Aalto Science Institute (AScI) internship programme

Summer 2021 project list

For more information on the program and how to apply, see
https://www.aalto.fi/en/aalto-science-institute/aalto-science-institute-asci-
internship-programme

SCHOOL OF SCIENCE 3

Department of Computer Science 3

1101 - Developing a DNA nanostructure design platform 3
1102 - Efficient visualisation and simulation of DNA strand-displacement systems 4
1103 - Statistical or psychological theories for combatting computer fatigue and multitasking 5
1104 - Statistical or psychological theories for user security 5
1105 - Active learning for interactive AI 6
1106 - Bayesian deep learning 6
1107 - Deep learning with differential equations 7
1108 - Training probabilistic circuits on large scale datasets 7
1109 - Towards Integrable Continuous Normalizing Flows 8
1110 - Non-stationary temporal priors for large-scale modelling 8
1111 - Efficient Model-based Path Integral Stochastic Control 10
1112 - Incorporating Physics into Deep Learning with Uncertainty 11
1113 - Deep Learning methods for Extreme Scale Classification 11
1114 - Modern ubiquitous applications: from devices to the cloud 12

Department of Neuroscience and Biomedical Engineering 13

1301 - Neuroimaging with naturalistic stimuli 13

Department of Applied Physics 14

1401 - Magnetic foams 14
1402 - Coarsening and flow of foams 14
1403 - Machine learning in statistical physics 14
1404 - Modeling and characterization of colored tandem solar cells 15
1405 - MBE growth of transition metal dichalcogenide heterostructures 17
1406 - Interaction-stabilized colossal topological Chern superconductors 17
1407 - Machine learning multipartite quantum entanglement 19
1408 - Quasiperiodic many-body criticality in quantum entangled magnets 20
1409 - Detection of entanglement in Cooper pair splitters 21
1410 - Quantum computing online 21
1411 - Single photon generation in electron quantum optics 23
1412 - Lee-Yang theory of Bose-Einstein condensation 24
1413 - Tunable couplers for superconducting qubits 25
1414 - FPGA programming of real-time feedback for quantum circuits 26
1415 - Open superconducting microwave cavity 27
1416 - Radiation damage in structural materials 28
1417 - Machine Learning Strategies for Scientific Data Analysis 29

P a g e 1 | 42
SCHOOL OF ELECTRICAL ENGINEERING 30

Department of Communications and Networking 30

2101 - Interactive prior elicitation 30
2102 - Bayesian Inference for Computational Cognitive Models 31
2103 - Helping People to Learn with Artificial Teachers that Adapt to User’s Cognitive Characteristics 32
2104 - AI-Assisted Indirect Input Design 33
2105 - Deep generative movement models for HCI 34
2106 - Computational methods in understanding decision-making processes in HCI 35
2107 - Interactive multi-objective optimization 36
2108 - Inferring agent preferences from behaviour 36
2109 - Model-based eye tracking 37
2110 - Computational Graphical User Interface Evaluation 37
2111 - Combining optimization with Machine learning for UI design 39
2112 - Interledger and Decentralised Identifiters for IoT Privacy 39

Department of Electronics and Nanoengineering 40

2201 - Organic Materials and Devices 40

SCHOOL OF CHEMICAL ENGINEERING 41

Department of Chemistry and Material Science 41

3101 - Investigation Li-ion batteries electrode degradation processes 41

SCHOOL OF ENGINEERING 42

Department of Civil Engineering 42

4101 - Computational analysis of microarchitectural solids and structures 42

P a g e 2 | 42
School of Science
Department of Computer Science

1101 - Developing a DNA nanostructure design platform

Field of study: Computer Science / Computational Biology

School: School of Science
Department: Computer Science
Professor: Pekka Orponen pekka.orponen@aalto.fi
Academic contact person: Pekka Orponen pekka.orponen@aalto.fi

The area of DNA nanotechnology [1] employs DNA as generic building material for assembling nanoscale
objects with dimensions in the order of 10-100 nanometres. Our group has been contributing, in collaboration
with a biochemistry team from Karolinska Institutet in Stockholm, to the design of a general-purpose design
platform “vHelix” for producing in particular 3D wireframe designs folded from a single long DNA strand [2].
A new, user-friendly and extendible version of the vHelix platform was developed as a summer internship
project in 2020 (not yet launched for public distribution). Following the publication of the DNA strand-routing
algorithm implemented in the current vHelix version in 2015 [3], several alternative methods have emerged,
and the goal of the present project is to implement some of these more recent algorithms as plugins to the new
extendible vHelix version.

The project requires familiarity with basic algorithm design techniques, facility with combinatorial thinking,
and good programming skills. Previous knowledge of biomolecules is not necessary, although it is an asset.

The work is performed in the context of research project “Algorithmic designs for biomolecular
nanostructures (ALBION)”, funded by the Academy of Finland. For further information, please see the
research group webpage at http://research.cs.aalto.fi/nc/.

[1] https://en.wikipedia.org/wiki/DNA_nanotechnology
[2] http://vhelix.net/
[3] Benson et al., Nature 2015, https://doi.org/10.1038/nature14586

P a g e 3 | 42
1102 - Efficient visualisation and simulation of DNA strand-displacement systems

Field of study: Computer Science / Computational Biology

School: School of Science
Department: Computer Science
Professor: Pekka Orponen pekka.orponen@aalto.fi
Academic contact person: Pekka Orponen pekka.orponen@aalto.fi

DNA strand-displacement (DSD) systems are a fundamental technique in the research area of dynamic DNA
nanotechnology [1], whose general goal is to implement dynamical behaviours, including computation, in
biomolecular systems. The basic biochemical DSD mechanism, which comprises the controlled binding and
release of short segments of single-stranded DNA, is quite simple, yet rich enough to support the design of
general-purpose computational devices [2].

The molecular-level details of DSD systems can get quite complicated and the eventual chemical
interactions rather intricate, and so high-level design, simulation and visualisation tools to support the
design of complex DSD systems are in demand. For this purpose, our group recently developed a flexible rule-
based modelling package RuleDSD, which at the simulation level integrates to the widely-used BioNetGen
framework [3]. A graphical user interface to the RuleDSD package, including a simulated annealing -based
strand network visualisation method, was developed as a summer internship project in 2020.

The first goal of the present project is to improve the efficiency and reliability of the RuleDSD strand
network visualisation tool, by replacing its general-purpose simulated annealing engine by an algorithmic
method designed specifically for this task. A second goal is to address the methodological and efficiency
challenges that arise in the current RuleDSD version when dealing with large, or even potentially infinite,
strand networks.

The project requires familiarity with basic algorithm design techniques, facility with combinatorial thinking,
and good programming skills. Previous knowledge of biomolecules is not necessary, although it is an asset.
The work is performed in the context of research project “Algorithmic designs for biomolecular
nanostructures (ALBION)”, funded by the Academy of Finland. For further information, please see the
research group webpage at http://research.cs.aalto.fi/nc.

[1] https://en.wikipedia.org/wiki/DNA_nanotechnology
[2] https://www.microsoft.com/en-us/research/project/programming-dna-circuits/
[3] Gautam et al., BIOSTEC 2020, https://doi.org/10.5220/0008979101580167

P a g e 4 | 42
1103 - Statistical or psychological theories for combatting computer fatigue and multitasking

Computer science, psychological science, social science,

Field of study:
behavioral science
School: School of Science
Department: Computer Science
Professor: Janne Lindqvist janne.lindqvist@aalto.fi
Academic contact person: Janne Lindqvist janne.lindqvist@aalto.fi

Aalto Human-Computer Interaction and Security Engineering Lab http://lindqvistlab.org at the Aalto
Department of Computer Science is recruiting a summer intern. We will be recruiting a student to contribute
to our research on the application of statistical theory or psychological theory to the study of combatting
computer fatigue and issues with multitasking. For example, we will study the cognitive mechanisms
responsible for computer fatigue and psychological principles related to issues with multitasking. Our
project will engage the summer intern in every step of the research process including research design,
statistical analysis and provide the intern with a unique exposure to the application of psychology and
statistics in human-computer interaction. Projects may also include Bayesian data analysis.

1104 - Statistical or psychological theories for user security

Computer science, psychological science, social science,

Field of study:
behavioral science
School: School of Science
Department: Computer Science
Professor: Janne Lindqvist janne.lindqvist@aalto.fi
Academic contact person: Janne Lindqvist janne.lindqvist@aalto.fi

Aalto Human-Computer Interaction and Security Engineering Lab http://lindqvistlab.org at the Aalto
Department of Computer Science is recruiting a summer intern. We will be recruiting a student to contribute
to our research on the application of statistical theory or psychological theory to the study of user security.
For example, we will study the cognitive mechanisms responsible for password usability and the
psychological principles driving security decisions. Our project will engage the summer intern in every step
of the research process including research design, statistical analysis and provide the intern with a unique
exposure to the application of psychology and statistics in human-computer interaction and usable security.
Projects may also include Bayesian data analysis. Accepted student also has the opportunity to participate to
activities of Helsinki-Aalto Institute for Cybersecurity (HAIC).

P a g e 5 | 42
1105 - Active learning for interactive AI

Field of study: Machine learning

School: School of Science
Department: Computer Science
Professor: Samuel Kaski samuel.kaski@aalto.fi
Academic contact person: Samuel Kaski samuel.kaski@aalto.fi

Most machine learning systems operate with us humans, to augment our skills and assist us in our tasks. In
environments containing human users, or, more generally, intelligent agents with specific goals and plans,
the system can only help them reach those goals if it understands them. Since the goals can be tacit and
changing, they need to be inferred from observations and interaction. We develop the probabilistic
interactive user models and inference techniques needed to understand other agents and how to assist them
more efficiently.

We are looking for a student to join us in developing these techniques. Additional keywords: active learning,
experimental design, knowledge elicitation, multiagent learning, machine teaching.

Link: https://aaltopml.github.io/machine-teaching-of-active-sequential-learners/

1106 - Bayesian deep learning

Field of study: Machine learning

School: School of Science
Department: Computer Science
Professor: Samuel Kaski samuel.kaski@aalto.fi
Academic contact person: Markus Heinonen markus.o.heinonen@aalto.fi

We are looking for a student to join the Aalto Probabilistic Machine Learning Group, to work on developing
state-of-the-art Bayesian deep learning. Key research questions are about more useful neural
parameterisations, process priors on function spaces and more efficient probabilistic inference methods for
deep neural networks. Possible applications range from large-scale image classification to sample-efficient
Bayesian reinforcement learning and robotics.

This work will build on top of existing research lines in the group on RL and BNNs, with a recent highlight
work of implicit BNNs with state-of-the-art ImageNet performance while maintaining Bayesian principles.
The group has excellent collaboration and application opportunities.

Background in machine learning, statistics or math is expected.

P a g e 6 | 42
1107 - Deep learning with differential equations

Field of study: Machine learning

School: School of Science
Department: Computer Science
Professor: Samuel Kaski samuel.kaski@aalto.fi
Academic contact person: Markus Heinonen markus.o.heinonen@aalto.fi

We are looking for an exceptional and motivated phd student to push the boundaries of deep learning with
differential equations. In conventional deep learning the inputs are transformed by a sequence of layers,
while an alternative paradigm emerged recently interpreting learning tasks as continuous flows with ODEs
or SDEs. We aim at developing new ways to perform machine learning by repurposing differential equations.

Possible topics include interpretable neural ODEs, to modelling data augmentations as flows, and to
probabilistic neural ODEs. The work is supported by multiple active research lines at the department.

Background with machine learning, statistics or math learning will be useful.

1108 - Training probabilistic circuits on large scale datasets

Field of study: Machine learning

School: School of Science
Department: Department of Computer Science
Professor: Arno Solin arno.solin@aalto.fi
Academic contact person: Martin Trapp martin.trapp@aalto.fi

Probabilistic circuits, such as sum-product networks and probabilistic sentential decision diagrams, are a
promising class of deep probabilistic models that unify ideas from classical AI and modern deep learning. In
contrast to many other probabilistic modelling families, probabilistic circuits guarantee exact and efficient
computation of many probabilistic inference tasks. However, training a circuit on massive datasets is a
challenging task due to their sparse architecture. The aim of this project is to adopt work on coresets to
enable efficient parameter learning in large scale settings.

The prospective candidates should be familiar with topics related to probabilistic machine learning and
should ideally have experience with python deep learning frameworks.

References:
[1] Peharz et al. (2020). Einsum Networks: Fast and Scalable Learning of Tractable Probabilistic Circuits, in
ICML. http://proceedings.mlr.press/v119/peharz20a/peharz20a.pdf
[2] Lucic et al. (2018). Training Gaussian Mixture Models at Scale via Coresets, in JMLR.
https://www.jmlr.org/papers/volume18/15-506/15-506.pdf

P a g e 7 | 42
1109 - Towards Integrable Continuous Normalizing Flows

Field of study: Machine learning

School: School of Science
Department: Department of Computer Science
Professor: Arno Solin arno.solin@aalto.fi
Academic contact person: Martin Trapp martin.trapp@aalto.fi

Normalizing flows use bijective, differentiable transformations to represent complex probability

distributions. Recent work has focused on so-called continuous normalizing flows, which utilize a
formulation in form of an ordinary differential equation (ODE) to represent more flexible flow architectures.
However, even though flow type models allow tractable computation of densities by construction they lack
the ability to compute more complex probabilistic inference tasks, e.g. marginals. In parallel, there has been
an increasing interest in probabilistic circuits, which use a much more restricted family of neural network
architectures to guarantee exact computation of arbitrary marginals, conditions and moments. The aim of
this project is to explore ways to utilize ideas from probabilistic circuits to enable exact computation of
marginals in continuous normalising flow models.

The prospective candidates should be familiar with topics related to probabilistic machine learning and
should ideally have experience with the Julia programming language or alternatively experience with Python
frameworks.

References:
[1] Pevny et al. (2020). Sum-Product-Transform Networks: Exploiting Symmetries using Invertible
Transformations, in PGM. https://pgm2020.cs.aau.dk/wp-content/uploads/2020/09/pevny20.pdf
[2] Z. Kolter, D. Duvenaud, and M. Johnson (2020). Deep Implicit Layers Tutorial at NeurIPS.
http://implicit-layers-tutorial.org

1110 - Non-stationary temporal priors for large-scale modelling

Field of study: Machine learning
School: School of Science
Department: Department of Computer Science
Professor: Arno Solin arno.solin@aalto.fi
Academic contact person: William Wilkinson william.wilkinson@aalto.fi

P a g e 8 | 42
We are looking for motivated, skilled, and open-minded summer students with an interest in real-time
inference and application of probabilistic machine learning methods to practical applications. This project is
concerned with non-stationary Gaussian process models and stochastic differential equations for temporal
and spatio-temporal data, with possible applications in air pollution modelling, sensor fusion, or computer
vision, where the problems are inherently noisy and uncertainty quantification plays an important role. An
important part of the project is implementing some of these models in JAX.

Successful applicants are expected to have an outstanding record in computer science, mathematics,
statistics, or a related field, and familiarity with some of the topics mentioned above.

References:
[1] Arno Solin (2020). Machine Learning with Signal Processing. ICML tutorial. Link:
https://youtu.be/vTRD03_yReI
[2] William J. Wilkinson, et al. (2020). State Space Expectation Propagation: Efficient Inference Schemes for
Temporal Gaussian Processes. In ICML. Pre-print: https://arxiv.org/abs/2007.05994

P a g e 9 | 42
1111 - Efficient Model-based Path Integral Stochastic Control

Field of study: Machine learning

School: School of Science
Department: Department of Computer Science
Professor: Arno Solin arno.solin@aalto.fi
Academic contact person: Vincent Adam vincent.adam@aalto.fi

Stochastic optimal control (SOC) is the problem of controlling a known noisy nonlinear dynamical system,
modeled by a stochastic differential equation (SDE) [1]. SOC has applications in many domains including
robotics and underlies many state-of-the-art reinforcement learning algorithms. When the system is
unknown or non-stationary, its statistics and parameters can be learned from experience. Gaussian
processes (GPs) can be used to incorporate a priori assumptions about the system (e.g., the drift of the SDE)
into the problem, leading to sample-efficient algorithms, i.e. algorithms that can achieve good performance
with a small number of interactions with the system. However, GPs suffer from poor computational scaling
making such algorithms either prohibitively slow and unfit to real world settings or require approximations
that impair their performance [2]. Recent advances in GP approximations now provide computationally
efficient ways to learn and sample from GPs with guarantees [3].

The aim of this project would be to combine the sample based Path-integral formulation of SOC and these
recent advances in GP inference and sampling to derive efficient model based SOC algorithms. The main aim
of the project would be to derive, implement and evaluate the resulting algorithm. If time allows, a more
theoretical analysis of the algorithm could be done.

The prospective candidates should be familiar with topics related to probabilistic machine learning and/or
reinforcement learning and should ideally have experience with Python.

References:
[1] Kappen (2008). Stochastic optimal control theory, ICML tutorial
[2] Pan et al. (2014). Model-based Path Integral Stochastic Control: A Bayesian Nonparametric Approach,
https://arxiv.org/abs/1412.3038
[3] Wilson et al. (2020). Pathwise Conditioning of Gaussian Processes, https://arxiv.org/abs/2011.04026

P a g e 10 | 42
1112 - Incorporating Physics into Deep Learning with Uncertainty

Field of study: Machine learning

School: School of Science
Department: Department of Computer Science
Professor: Arno Solin arno.solin@aalto.fi
Academic contact person: Vincent Adam vincent.adam@aalto.fi

Dynamical systems are governed by the laws of physics. Learning a dynamical system supporting accurate
long-term predictions from observations of trajectories can be greatly enhanced by incorporating the right
inductive biases from physics (e.g., in the form of constraints and conservation laws). Recent work has
successfully incorporated laws of physics into deep learning [1, 2]. For example, neural networks were used
to parameterize the parameters of the Hamiltonian of a system and learning was achieved through back
propagating through the solution of the equations of motion derived from this Hamiltonian [3]. The purpose
of this project would be to investigate how to best incorporate uncertainty about parameters in such models.
Parameter uncertainty leads to uncertain future predictions which are crucial in control scenarios where
driving the system in some particular configurations is very costly.

The prospective candidates should be familiar with topics related to classical physics, probabilistic machine
learning or deep learning and should ideally have experience with deep learning frameworks such as
TensorFlow.

References:
[1] Lutter et al. (2020). Deep Lagrangian Networks: Using Physics as Model Prior for Deep Learning,
https://arxiv.org/abs/1907.04490
[2] Steindor Saemundsson et al. (2020). Variational Integrator Networks for Physically Structured
Embeddings, ICML.
[3] Chen et al. (2018). Neural Ordinary Differential Equations, NeurIPS.

1113 - Deep Learning methods for Extreme Scale Classification

Field of study: Machine Learning

School: School of Science
Department: Department of Computer Science
Professor: Rohit Babbar rohit.babbar@aalto.fi
Academic contact person: Rohit Babbar rohit.babbar@aalto.fi

Large output spaces with hundreds of thousand labels are common in Machine learning problems such as
ranking, recommendation systems and next word prediction. Apart from the computational problem of
scalability, data scarcity for individual labels poses a statistical challenge and especially so for data hungry
deep methods. The goal of the project is to investigate deep learning based architectures and adapting the well
known techniques such as Attention mechanism for simultaneously addressing the computational and
statistical challenge in learning with large output spaces. As the target domain is textual data, the project also
involves exploring recent advances in NLP, such as Bert and TransformerXL, towards exploring the common
grounds for further research in this area.

P a g e 11 | 42
1114 - Modern ubiquitous applications: from devices to the cloud

Field of study: Computer Science

School: School of Science
Department: Department of Computer Science
Professor: Mario Di Francesco mario.di.francesco@aalto.fi
Academic contact person: Mario Di Francesco https://users.aalto.fi/~difram1/

Modern applications that are ubiquitous – namely, everywhere – rely on two key components. First, on
embedded devices such as mobile phones, wearables and smart objects in the Internet of Things to interact
with users and collect information from the physical environment. Second, on a cloud or edge computing
infrastructure to support different types of applications requiring a substantial amount of processing and
storage, such as those involving machine learning. The major challenges in realizing such applications
include efficient resource utilization at both devices and the supporting infrastructure, reliability, and user
friendliness. The goal of this project is to investigate some of these aspects in the context of the research
carried out in our research group. See also https://users.aalto.fi/~difram1/ for additional details.

Required skills: experience with Android application development or embedded systems programming,
solid understanding of data analysis and (or) machine learning techniques.

Desired skills: some knowledge on human computer interaction and (or) computer vision, familiarity with
cloud and web technologies.

P a g e 12 | 42
Department of Neuroscience and Biomedical Engineering

1301 - Neuroimaging with naturalistic stimuli

Field of study: Neuroimaging, naturalistic stimuli, data analysis

School: School of Science
Department: Department of Neuroscience and Biomedical Engineering
Professor: Iiro Jääskeläinen Iiro.jaaskelainen@aalto.fi
Academic contact person: Iiro Jääskeläinen Iiro.jaaskelainen@aalto.fi

The topic of the position is functional magnetic resonance imaging (fMRI) using naturalistic stimuli such as
movies and narratives. Using naturalistic stimuli has opened possibilities especially to address questions on,
e.g., the neural basis of emotions and social cognition, that would have been difficult to tackle well with a more
traditional experimental approach. This approach has especially flourished with advances in data analysis
methods, such as intersubject correlation and machine learning approaches. The summer internship would
involve the prospective student with aspects of experimental design with naturalistic stimuli, acquisition of
data, and data analysis methods.

Prerequisites to the position include keen interest in the area, basic knowledge of cognitive neuroscience (e.g.,
some courses completed), and general skills in data analysis (e.g., basics of statistics, machine learning).
Familiarity with Matlab and/or Python environments is a considered advantageous.

P a g e 13 | 42
Department of Applied Physics

1401 - Magnetic foams

Field of study: Physics

School: School of Science
Department: Department of Applied Physics
Professor: Mikko Alava mikko.alava@aalto.fi
Academic contact person: Mikko Alava mikko.alava@aalto.fi

Here we study complex systems in the shape of magnetic foams. These are made of foams where magnetic
nanoparticles have been mixed in. The intern is supposed to be interested in the physics of soft matter, and
have skills or be interested in learning experimental physics and data analysis. The project is a collaboration
between two groups at the department (Alava and ERC Grant winner Jaakko Timonen) and involves
experimental studies of such foams. The project continues the successful topic of self-propelled particles in
foams from summer 2019.

1402 - Coarsening and flow of foams

Field of study: Physics

School: School of Science
Department: Department of Applied Physics
Professor: Mikko Alava mikko.alava@aalto.fi
Academic contact person: Antti Puisto antti.puisto@aalto.fi

Foams coarsen i.e. change their structure with time. This is a complex process involving several microscopic
mechanisms forcing the topology of the foam to change accordingly. Since, external forcing influences the
foam structure, the coarsening in flowing foam systems is expected to deviate heavily from that in static
foams. Here, we attack this problem by 3D simulations of foams using a novel GPU- based simulation code.
The candidates should have an interest in soft matter physics and/or computational physics.

1403 - Machine learning in statistical physics

Field of study: Physics

School: School of Science
Department: Department of Applied Physics
Professor: Mikko Alava mikko.alava@aalto.fi
Academic contact person: Mikko Alava mikko.alava@aalto.fi

P a g e 14 | 42
The task is to utilize the recently developed machine learning and artificial intelligence algorithms in statistical
physics problems. The previous topics include predicting catastrophic events and local yielding or
automatically classifying the material properties. Here, the algorithms are adapted to similar existing
experimental and numerical datasets. The candidate is expected to have some programming experience and
keen interest in computational physics.

1404 - Modeling and characterization of colored tandem solar cells

Field of study: Applied physics, Photonics, Photovoltaics

School: School of Science
Department: Department of Applied Physics
Professor: Janne Halme janne.halme@aalto.fi
Janne Halme janne.halme@aalto.fi
Academic contact person:
Farid Elsehrawy farid.elsehrawy@aalto.fi

Description: Building-integrated photovoltaics allow using a larger area for integration of solar panels into
building façades – the visible part of the building envelope. However, conventional black solar panels do not
appeal to architects and users, who prefer lighter colors for building façades [1]. Traditional colored solar cells
have limited power that can be harvested due to the optical losses associated with colors [2]. Accordingly,
colored solar cells can be implemented using tandem solar cells, which are less sensitive to color-induced
optical losses than conventional silicon solar cells due to their efficient utilization of the sunlight spectrum.
Tandem solar cells increase the efficiency by stacking multiple junctions to harvest light from a wider solar
spectral range. The optical/electronic modeling and optimization of colored tandem solar cells requires a
comprehensive approach involving the dependence of the optimal tandem solar cell bandgaps on the color
spectrum. In this project, we develop colored solar panels that are visually aesthetic and efficient. To maximize
efficiency, we search for the optimal coloration process for tandem solar cells.

References
[1] Bao et al., Renew. Energy, 113 (2017).
[2] Halme et al., Energy Environ. Sci., 12, 4 (2019).

Role and duties: The candidate’s role will be to develop a modeling tool for colored tandem solar cells that
enables optical simulations of the color coatings in addition to optical/electronic simulations of tandem solar
cells. Next, the candidate will validate the functionality of the tool by performing optical and electrical
characterization of colored tandem solar cells.

Team: The work will be performed within the New Energy Technologies research group at the Department of
Applied Physics, Aalto University. The colored photovoltaics research area is led by Dr. Janne Halme,
University Lecturer. The candidate will work within a team of experts on colored photovoltaics and opto-
electronic solar cell modeling and characterization.

Necessary skills: Strong background in optics, numerical modeling, and solid understanding of solar cells
operation. Experience in programming using Matlab and/or Python.

P a g e 15 | 42
P a g e 16 | 42
1405 - MBE growth of transition metal dichalcogenide heterostructures

Field of study: Phyics

School: School of Science
Department: Department of Applied Physics
Professor: Peter Liljeroth peter.liljeroth@aalto.fi
Academic contact person: Xin Huang xin.huang@aalto.fi

This project involves growing heterostructures of transition metal dichalcogenides using molecular-beam
epitaxy (MBE) and characterizing them using surface science methods (XPS, STM, AFM) and optical
techniques (e.g. micro-Raman). The main part of the work is to optimize the MBE growth parameters for
obtaining well-defined heterostructures.

The applicant should have some experience on working a lab, prior experience on MBE or surface science
tools is a plus.

1406 - Interaction-stabilized colossal topological Chern superconductors

Theoretical physics, Topological physics,

Field of study:
Interplay between magnetism and superconductivity
School: School of Science
Department: Department of Applied Physics
Professor: Jose Lado jose.lado@aalto.fi
Jose Lado jose.lado@aalto.fi
Academic contact person: Timo Hyart timo.hyart@aalto.fi
Teemu Ojanen teemu.ojanen@tuni.fi

P a g e 17 | 42
Topological Chern superconductors are one of the paradigmatic milestones in quantum materials, due to the
emergence of Majorana modes associated with their topological electronic structure. Ultimately, topological
superconductors represent one of the critical materials for potential future topological quantum computers,
stemming from the non-abelian braiding properties of their topological excitations. Topological
superconductors are extremely rare in nature, which has motivated intense theoretical and experimental
efforts for their design in artificial materials. Remarkably, recent breakthrough experiments have
demonstrated the emergence of Majorana modes in these compounds. Theoretically, these systems can
support an arbitrarily large number of topological modes. However, experiments up to date only have realized
topological superconducting states with a limited number of Majorana modes.

The objective of this project is to theoretically study a model for topological superconductor including
collective many-body reconstructions driven by electronic interactions. In particular, the student will
demonstrate how electronic interactions can dramatically enhance topological gaps of interacting Chern
superconductors, stabilizing phases displaying a large unprecedented number of Majorana modes. During the
internship, the student will gain in-depth knowledge of topological physics, superconductivity, and
magnetism, providing him/her with an outstanding experience for future studies in theoretical physics and
quantum materials.

Necessary skills:
We look for a highly motivated BSc/MSc student in Physics, with a strong background in condensed matter
physics, theoretical physics, and computational physics. The project will combine several analytic and
computational skills, and as a result, we look for a student with experience in programming, ideally with
Python.

P a g e 18 | 42
1407 - Machine learning multipartite quantum entanglement

Theoretical physics
Field of study: Many-body quantum physics
Machine learning
School: School of Science
Department: Department of Applied Physics
Professor: Jose Lado jose.lado@aalto.fi
Academic contact person: Jose Lado jose.lado@aalto.fi

Quantum many-body systems represent one of the cornerstones of modern physics, providing platforms to
explore radical new forms of exotic physical phenomena. This genuine emergent phenomena in many-body
systems stems from the existence of quantum entanglement, non-local forms of correlation that are forbidden
in classical systems. However, experimentally detecting generic forms of quantum entanglement has
remained an open problem, as only partial information of a quantum state can be accessed in experiments.
Machine learning raised in the last years as a compelling framework for extracting highly-non trivial
information from physical systems. Ultimately, machine learning algorithms have demonstrated to provide a
powerful formalism to solve complex physical problems that lack a conventional solution, and even in the
presence of incomplete and noisy information.

The objective of this project is to demonstrate that a machine-learning algorithm to infer multipartite
entanglement from measured correlation functions. In particular, the student will demonstrate how
supervised machine learning with deep neural networks predicts entanglement properties of quantum-many
body systems solved with tensor networks. During the internship, the student will gain in-depth knowledge of
machine learning, quantum entanglement, and many-body physics, providing him/her with an outstanding
experience for future studies in theoretical physics, quantum technologies, and quantum materials.

Necessary skills:
We look for a highly motivated BSc/MSc student in Physics, with a strong background in condensed matter
physics, theoretical physics, and computational physics. The project will combine analytic and computational
skills, and so we look for a student with experience in programming, ideally with Python.

P a g e 19 | 42
1408 - Quasiperiodic many-body criticality in quantum entangled magnets

Theoretical physics
Field of study: Many-body quantum physics
Quasiperiodic criticality
School: School of Science
Department: Department of Applied Physics
Professor: Jose Lado jose.lado@aalto.fi
Academic contact person: Jose Lado jose.lado@aalto.fi

Interacting quantum systems represent one of the paradigmatic open problems in modern condensed matter
physics. In particular, these systems are known to display remarkable emergent phenomena, including
macroscopic quantum entanglement and exotic fractionalized excitations. Besides the conventional periodic
quantum systems that are realized in natural materials, the rise of metamaterials and moire heterostructures
has put forward a whole new class of systems lacking conventional translational symmetry but rather
realizing quasiperiodic order. Quasiperiodicity is known to give rise to a whole new family of unconventional
phenomena, and in particular, leading to the realization of tunable quantum criticality and high dimensional
topological phenomena.

The objective of this project is to theoretically study a model for a quasiperiodic many-body one-dimensional
quantum magnet using the many-body tensor network formalism. In particular, the student will demonstrate
how the coexistence of quantum entanglement and quasiperiodicity leads to the emergence of critical
fractionalized spinon excitations. During the internship, the student will gain in-depth knowledge of
quasiperiodic physics, fractionalization, and quantum magnetism, providing him/her with an outstanding
experience for future studies in theoretical physics, quantum technologies, and quantum materials.

Necessary skills
We look for a highly motivated BSc/MSc student in Physics, with a strong background in condensed matter
physics, theoretical physics, and computational physics. The project will combine analytic and computational
skills, and so we look for a student with experience in programming, ideally with Python.

P a g e 20 | 42
1409 - Detection of entanglement in Cooper pair splitters

Theoretical physics,
Field of study:
Quantum transport, Quantum information
School: School of Science
Department: Department of Applied Physics
Professor: Christian Flindt christian.flindt@aalto.fi
Academic contact person: Fredrik Brange fredrik.brange@aalto.fi

Summary
Entanglement lies at the heart of quantum mechanics and constitutes a key ingredient for many quantum
information applications, such as quantum teleportation, superdense coding, and quantum cryptography.
Over the last two decades, Cooper pair splitters have emerged as a promising way of generating entangled
electrons in nanoscale systems. In these systems, entangled pairs of electrons are extracted from a
superconductor and spatially separated to, e.g., two quantum dots. In this project, we will investigate how
entanglement can be detected in a quantum dot-based Cooper pair splitter, and devise an experimental
scheme to verify (and quantify) the entanglement of the emitted electrons by using entanglement witnesses, a
general method for detecting entanglement. The student will gain in-depth knowledge in how to describe
transport and entanglement in quantum systems, providing him/her with a solid background for future
studies and research in theoretical physics, quantum transport and quantum information.

Necessary skills
We look for a highly motivated student in theoretical physics (or related areas), with a strong academic
background in quantum physics and/or mathematics. The project will mainly deal with analytic derivations
and calculations.

1410 - Quantum computing online

Theoretical physics, Quantum information, Quantum

Field of study:
computing
School: School of Science
Department: Department of Applied Physics
Professor: Christian Flindt christian.flindt@aalto.fi
Academic contact person: Fredrik Brange fredrik.brange@aalto.fi

P a g e 21 | 42
Summary
The field of quantum information and quantum computing is currently undergoing a tremendous
development. Not long ago, only very few qubits had been realized experimentally in the lab. Now, small
versions of real quantum computers are being developed and, in some cases, made accessible online by a few
different companies. In this project, we will implement quantum computing algorithms remotely on one of
these real quantum computers. You will have the chance to get familiar with both the theory of quantum
information and quantum computing as well as the practical aspects of using quantum computers remotely.
After the project, you will have gained a solid background for future studies and research in theoretical
physics, quantum information and quantum computing.

Necessary skills
We look for a highly motivated student in theoretical physics (or related areas), with a strong academic
background in quantum physics and programming.

P a g e 22 | 42
1411 - Single photon generation in electron quantum optics

Field of study: Theoretical physics, quantum mechanics, quantum optics

School: School of Science
Department: Department of Applied Physics
Professor: Christian Flindt christian.flindt@aalto.fi
Academic contact person: Benjamin Roussel benjamin.roussel@aalto.fi

Summary
Recent experiments have demonstrated the ability to generate, manipulate, and probe electrical currents at the
level of the individual electrons. This progress has led to a novel research field in condensed matter physics
called electron quantum optics, where electrons are manipulated in a similar way as photons in quantum optics.
However, there are major differences between photons in a waveguide and electrons in a quantum conductor:
Electrons, being charged particles, are always surrounded by electromagnetic radiation. Notably, this means
that it is possible to use electron quantum optics to engineering interesting states of the electromagnetic field
by controlling the electronic states with time-dependent voltages. In this theoretical project, the student will
explore the implementation of a single-photon source based on the decay of electron-hole pairs in a quantum
conductor. The student will gain in-depth knowledge of quantum optics and interactions in many-body
electronic systems, providing her/him with a solid expertise for further studies in theoretical physics and
quantum technologies.

Necessary skills
We are looking for a highly motivated student in theoretical physics (or related areas), with a strong
knowledge of quantum mechanics. This project will mainly deal with analytical derivations and simple
numerical calculations.

P a g e 23 | 42
1412 - Lee-Yang theory of Bose-Einstein condensation

Field of study: Theoretical physics, quantum mechanics, statistical physics

School: School of Science
Department: Department of Applied Physics
Professor: Christian Flindt christian.flindt@aalto.fi
Academic contact person: Christian Flindt christian.flindt@aalto.fi

Summary:
Bose-Einstein condensation occurs when a gas of bosons is cooled below its critical temperature, and a large
fraction of the particles suddenly occupy the ground state of the system. Bose-Einstein condensation is an
example of a phase transition, where the macroscopic properties of a many-body system abruptly changes as an
external control parameter is varied. In two seminal papers, Lee and Yang showed that phase transitions can be
understood by considering the zeros of the partition function in the complex plane of the control parameter: In
the thermodynamic limit of large system sizes, the zeros will approach the real value of the control parameter
for which a phase transition occurs, and the associated free energy as well as other thermodynamic observables
become nonanalytic. For a long time, these Lee-Yang zeros were considered a purely theoretical concept.
Recently, however, it has been realized that they can also be experimentally determined, leading to a renewed
interest in Lee-Yang theories.

In this project, the student will make use of recent advances in the theoretical understanding of phase
transitions in many-body systems and employ a cumulant method to determine the Lee-Yang zeros of Bose-
Einstein condensation from the energy fluctuations in small Bose gases. Using a finite-size scaling analysis, it
is possible to extrapolate the convergence points of the Lee-Yang zeros in the thermodynamic limit and thereby
determine the condensation temperature, for instance, depending on the dimensionality and the shape of the
trapping potential. The project is theoretical but will lead to specific suggestions for future experiments on Bose
gases. The student will become acquainted with recent theories of phase transitions in quantum many-body
systems, which will provide him/her with a solid background for future studies and research in theoretical
quantum and statistical physics.

Necessary skills:
We are looking for a highly motivated student in theoretical physics (or related areas), with a strong academic
background in quantum physics, statistical mechanics and/or mathematics. The project will mainly deal with
analytic derivations and calculations as well as some numerics. Experience with Mathematica, for instance, is
an advantage.

P a g e 24 | 42
1413 - Tunable couplers for superconducting qubits

Experimental quantum physics

Field of study:
Superconducting qubit - mechanical hybrid quantum systems
School: School of Science
Department: Department of Applied Physics
Professor: Mika A. Sillanpää Mika.Sillanpaa@aalto.fi
Academic contact person: Mika A. Sillanpää Mika.Sillanpaa@aalto.fi

In the Quantum Nanomechanics research group

https://www.aalto.fi/en/department-of-applied-physics/quantum-nanomechanics
we work on nanomechanical systems in the quantum regime, and on superconducting quantum hybrid
systems. Experimental work is carried out in the premises of Low Temperature Laboratory.

The summer projects involve design, fabrication and measurement of devices, and give an excellent overview
of cutting-edge experimental research on an exciting topic with a strong relevance to quantum technologies.
Quantum bits made with Josephson junctions are considered the most promising platform for realization of
quantum computer. Besides this distant goal, superconducting qubits can be useful for exploring hybrid
quantum systems, and testing quantum mechanics in nearly macroscopic systems. To this end, we offer three
summer projects:

1. Tunable couplers for superconducting qubits.

In a quantum computation architecture, one has to be able to switch on or off the interaction between adjacent
qubits. In this project, you will design a fast switchable coupling circuit for two transmon qubits. You will first
run electromagnetic simulations to design the structure, then realize the device using cleanroom
microfabrication, and finally test the device in a dilution refrigerator.

P a g e 25 | 42
1414 - FPGA programming of real-time feedback for quantum circuits

Experimental quantum physics

Field of study:
Superconducting qubit - mechanical hybrid quantum systems
School: School of Science
Department: Department of Applied Physics
Professor: Mika A. Sillanpää Mika.Sillanpaa@aalto.fi
Academic contact person: Mika A. Sillanpää Mika.Sillanpaa@aalto.fi

In the Quantum Nanomechanics research group

https://www.aalto.fi/en/department-of-applied-physics/quantum-nanomechanics
we work on nanomechanical systems in the quantum regime, and on superconducting quantum hybrid
systems. Experimental work is carried out in the premises of Low Temperature Laboratory.

The summer projects involve design, fabrication and measurement of devices, and give an excellent overview
of cutting-edge experimental research on an exciting topic with a strong relevance to quantum technologies.
Quantum bits made with Josephson junctions are considered the most promising platform for realization of
quantum computer. Besides this distant goal, superconducting qubits can be useful for exploring hybrid
quantum systems, and testing quantum mechanics in nearly macroscopic systems. To this end, we offer three
summer projects:

2. FPGA programming of real-time feedback for quantum circuits

Superconducting qubits, or quantum micromechanical experiments, often require real-time feedback, built
upon the measurement record, in order to apply delicate corrections to the quantum state. In this project you
will program a FPGA (Field Programmable Gate Array) data processing architecture for creating such
feedback. The system is tested with simulated data, and depending on the progress, also tests using real
samples can be foreseen.

P a g e 26 | 42
1415 - Open superconducting microwave cavity

Experimental quantum physics

Field of study:
Superconducting qubit - mechanical hybrid quantum systems
School: School of Science
Department: Department of Applied Physics
Professor: Mika A. Sillanpää Mika.Sillanpaa@aalto.fi
Academic contact person: Mika A. Sillanpää Mika.Sillanpaa@aalto.fi

In the Quantum Nanomechanics research group

https://www.aalto.fi/en/department-of-applied-physics/quantum-nanomechanics
we work on nanomechanical systems in the quantum regime, and on superconducting quantum hybrid
systems. Experimental work is carried out in the premises of Low Temperature Laboratory.

The summer projects involve design, fabrication and measurement of devices, and give an excellent overview
of cutting-edge experimental research on an exciting topic with a strong relevance to quantum technologies.
Quantum bits made with Josephson junctions are considered the most promising platform for realization of
quantum computer. Besides this distant goal, superconducting qubits can be useful for exploring hybrid
quantum systems, and testing quantum mechanics in nearly macroscopic systems. To this end, we offer three
summer projects:

3. Open superconducting microwave cavity

Three-dimensional (3D) microwave cavity resonators offer an attractive platform for implementing the best
superconducting qubit circuits. Additionally, 3D cavities can be used to carry out quantum experiments with
large vibrating membranes used as mechanical oscillators. In this project, you will investigate a possibility to
realize an open 3D cavity that consists of two separate blocks. The motivation is that the scheme allows for
introducing voltage to one of the blocks. The work involves running electromagnetic simulations, realizing the
physical structure, and carrying out basic measurements.

P a g e 27 | 42
1416 - Radiation damage in structural materials

Field of study: Physics

School: School of Science
Department: Department of Applied Physics
Professor: Andrea Sand andrea.sand@aalto.fi
Academic contact person: Andrea Sand andrea.sand@aalto.fi

Description: Particle irradiation modifies the physical and mechanical properties of materials, and plays an
increasing role in modern technological developments. For example, climate change is driving the need for
green energy, with nuclear fusion and next generation fission standing as two of the strongest candidates for
efficient and reliable energy production of the future, yet the challenges posed to reactor materials in the high
radiation environments are significant. Modelling provides an essential tool for predicting the response of
reactor components in future nuclear devices. The damage in materials created by energetic impacting
particles is highly sensitive to the mechanisms of dissipation of the impinging particle’s kinetic energy. This
summer project involves performing simulations employing a recently developed atomistic model, which
accounts for energy dissipation in unprecedented detail, to predict the primary radiation damage in metals
and model alloy systems under different incident neutron and ion energy spectra. The work will support the
further development and validation of the model. The student will gain knowledge of the processes of
radiation damage formation in materials, learn the basics of performing molecular dynamics simulations of
highly non-equilibrium events, and develop a familiarity with high performance computing environments.

Necessary skills: Experience in programming is highly desirable. The candidate should also have basic
knowledge of solid state physics and computational physics. Previous experience of molecular dynamics or
high performance computing is considered a plus.

P a g e 28 | 42
1417 - Machine Learning Strategies for Scientific Data Analysis

Field of study: Physics

School: School of Science
Department: Department of Applied Physics
Professor: Adam Foster adam.foster@aalto.fi
Academic contact person: Adam Foster adam.foster@aalto.fi

Scientific data can be generated through physical simulations, experimental laboratories and
observations from real-world problems. Compared to just a few years ago, the advancement of scientific
instruments, digital sensors and computational resources as well as storage devices have created huge
collections of scientific data. Unlike traditional statistical analysis, Machine Learning (ML) thrives on
growing data sets. The more data fed into an ML system, the more it can learn and apply the results to
higher quality predictions and new insights. In this project, we will investigate and implement ML
methods (e.g., kernel regression, autoencoders, deep learning) for finding key variables influencing
physical phenomena and materials properties. In particular, we will develop and exploit the wealth of
materials data available (most of it generated in our research group), and use ML to discover new
materials and phenomena linked to them. Examples within the SIN group (http://www.aalto.fi/physics-
sin) include interpreting microscopy imaging, identifying exotic quantum phenomena and predicting
hydration structures.

The detailed applications and tasks will be tailored according to the background of successful candidates.
Applicants should have a basic knowledge of physics, data analysis and statistics. Knowledge of Python would
be highly beneficial.

P a g e 29 | 42
School of Electrical Engineering
Department of Communications and Networking

2101 - Interactive prior elicitation

Computer Science - HCI, Probabilistic Machine Learning,

Field of study:
Cognitive Science, Bayesian Methods
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Suyog Chandramouli suyog.chandramouli@aalto.fi

Description: Prior elicitation, during which subjective knowledge and beliefs of experts are elicited is an
important part of the Bayesian methodology which is used very widely in artificial intelligence and
computational modeling. However, there are many sources of variance in elicited beliefs due to (i) actual
differences in experts' beliefs, (ii) cognitive biases of the experts in uncertain settings, (iii) measurement noise
in prior elicitation methods, (iv) elicitor induced biases, and (v) methodological flexibility in the elicitation
procedure. The goal of any good prior elicitation method is to capture to the extent possible, the true unbiased
beliefs of an expert while making sensitive measurements.
Prior elicitation methods are often developed independently, and seldom comprehensively tested for validity.
Additionally, variance in elicited beliefs are often assumed to be only due to differing views of experts. In this
project we will develop a new interactive prior elicitation procedure which views human judgments as a result
of sampling from subjective mental representations, after a training phase using data generating distributions
that are known to us. This procedure which collects many such samples from participants will then be used as
a gold standard along with the true generating distributions, against which existing elicitation methods (which
we believe may be quicker but less accurate) would also be compared. [This work is associated with the Finnish
Center for AI (fcai.fi).]

Preferred Skills/qualification
Experience with one or more of the following:
a) Bayesian statistical methods, e.g. MCMC sampling
b) Programming HCI or Psychology experiments, e.g. using PsychoPy
c) Cognitive modeling
d) R, Python

P a g e 30 | 42
2102 - Bayesian Inference for Computational Cognitive Models

Computer Science - Human-Computer Interaction, Machine

Field of study:
Learning, and Cognitive Science
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Antti Oulasvirta antti.oulasvirta@aalto.fi

Description: Computational cognitive models are simulators that make predictions of behaviour in various
tasks. They can be very useful in creating intelligent interactive systems that understand their users, but only
if the models are correctly specified and their parameters set according to the preferences and abilities of
individual users. Correctly parametrised user models can predict behaviour under various conditions,
permitting optimisation of user interfaces and anticipation of user responses to task events. Recent advances
in reinforcement learning (RL) based cognitive modelling have extended the range of possibilities for
simulating task behaviour, as they can predict how adaptive strategies emerge in various task conditions.

The project presents an opportunity to learn about RL based computational modelling and Bayesian
parameter inference. It utilises an existing RL based computational model and task data, and utilises an
engine for inferring the parameters of the model from different individuals in the data. The project
implements the inference framework required to create predictions of users given observed interactive
behaviour. [This work is associated with the Finnish Center for AI (fcai.fi).]

Requirements: The applicants are expected to have strong programming skills (Python) and studies in
human-computer interaction. In addition, skills in following are preferred but not required: reinforcement
learning, Bayesian parameter inference, computational cognitive modelling.

P a g e 31 | 42
2103 - Helping People to Learn with Artificial Teachers that Adapt to User’s Cognitive
Characteristics

Computer Science - Human-Computer Interaction, Machine

Field of study:
Learning, and Cognitive Science
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Aurélien Nioche aurelien.nioche@aalto.fi

Description: How can we make better artificial teachers? We suggest splitting the problem into two parts: (i)
inference of the cognitive features, (ii) sequence planning. Hence, on the one hand, we infer the cognitive
traits of the user, by fitting the user data with a series of plausible psychological models. On the other hand, the
schedule of review is optimized based on individualized predictions of performance. This generic framework
allows answering a series of questions that would help to develop better self-teaching applications: How to
take into account the cognitive particularities of the learner? How to take into account the time constraints of
the learner? How to combine the information of several learners? How to maximize the information about the
learner without penalizing his learning (i.e. how to deal with the exploitation-exploration tradeoff in this
context)? How to select the ‘teaching’ trajectories worthy to simulate (i.e. identify the best sampling technique
in this context)? This project includes in silico experiments (with artificial agents only) and lab experiments
(with human embodied users). The intern will aim to improve user modeling and/or the algorithm that
handles the sequence planning. The intern will be invited to also test his own hypothesis. [ This work is
associated with the Finnish Center for AI (fcai.fi).]

Requirements: The intern is expected to know to be proficient in Python. Having some background
knowledge about data analysis, machine learning, and/or cognitive modeling will be appreciated.

P a g e 32 | 42
2104 - AI-Assisted Indirect Input Design

Computer Science - Human-Computer Interaction, Machine

Field of study:
Learning
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Julien Gori julien.gori@aalto.fi

Description: Transfer functions in indirect input modalities operate a transformation from physical action to
virtual movement (e.g. of a cursor on a screen), and are usually designed by trial-and-error.
The goal of this project is to propose a general methodology for designing transfer functions, that takes into
account explicit objectives (e.g. minimizing fatigue of a user) and users own abilities (e.g. motor impaired
user). We first look at this problem under the scope of a two agent paradigm. The first agent, the operator,
provides physical actions (e.g. mouse movements), which are modulated by a second agent, the enhancer (e.g.
applying a constant gain). The operator wants to reach a specific goal (e.g. a target on the screen): the enhancer
will help him reach that goal in the best possible way i.e. accounting for the operators abilities and the stated
objective.
The behavior of the two agents will be determined using the multi-agent reinforcement learning (MARL)
framework, "in silico". A second step will be to deploy the enhancer with a human operator. The enhancer's
performance will be validated with several devices and with different operators. The project thus presents the
opportunity for the student 1) to develop his skills in reinforcement learning (RL), including MARL; 2) to
implement a working solution on a chosen indirect input device. [This work is associated with the Finnish
Center for AI (fcai.fi).]

Preferred Skills/qualification: Strong programming skills (Python, GUI) are expected.

P a g e 33 | 42
2105 - Deep generative movement models for HCI

Computer Science - Human-Computer Interaction, Machine

Field of study:
Learning
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Julien Gori julien.gori@aalto.fi

Description: This project will be in collaboration with Prof. Luis Leiva (University of Luxemburg).
Movements, from cursor trajectories, to strokes, gestures, and handwriting are ubiquitous in Human
Computer Interaction (HCI). Artificially generating human-like movement data would allow to
improve the design of HCI interfaces while reducing the need for time consuming experiments with
participants in the laboratory.
The articulation and production of human movements is a complex problem grounded on motor
control principles. Modern deep generative models can produce realistic human-like movements,
exploiting non-linear and long-term dependencies within the observed human data. However, these
models can make only limited inferences about the distribution of the observed data.
The project will focus on modeling the full representation of human handwriting, gestures and/or
pointing, i.e. 2D trajectories with their associated timestamps. The intern will contribute techniques
to sample realistic trajectories from deep generative models. [This work is associated with the Finnish Center
for AI (fcai.fi).]

Preferred Skills/qualification: The intern is expected to be knowledgeable in Deep Learning (DL) methods
for generative modeling and be familiarized with some popular DL library or framework, such as Tensorflow,
Keras, Pytorch, Lightning, etc.

P a g e 34 | 42
2106 - Computational methods in understanding decision-making processes in HCI

Computer Science - Human-Computer Interaction, Machine

Field of study:
Learning, and Cognitive Science
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Aini Putkonen aini.putkonen@aalto.fi

Description: People make hundreds, if not thousands, of decisions a day. These decisions reflect individual
preferences, resulting from latent cognitive and behavioural processes. Today, individuals make a number of
these decisions when interacting with different technologies, clicking and tapping through content on their
devices. Understanding the rationale behind each click and tap can be leveraged in various applications in
human computer interaction, including personalising user interfaces. Drawing from existing research in
decision theory and advances in computational cognitive modelling, we aim to discover novel ways of
understanding decision-making using empirical data. The types of questions we aim to answer include: Which
models best capture how individuals engage with content, e.g. news articles, on a website? How can we model
users’ scrolling behaviour? What kind of data can be used to test these models?
The intern will contribute to both model specification and practical implementation of the modelling
workflow, with an emphasis depending on their specific interest. We invite candidates with a background in
social/cognitive/computer science, and other applicable fields. An ideal candidate is expected to be proficient
in Python and have an understanding of some model types used in HCI and cognitive science (e.g. POMDP,
heuristics, models of decision-making under risk). [This work is associated with the Finnish Center for AI
(fcai.fi).]

Requirements: Experience in handling and analysing large datasets is appreciated.

P a g e 35 | 42
2107 - Interactive multi-objective optimization

Computer Science - Mathematics, Machine learning,

Field of study:
Optimization Techniques
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Morteza Shiripour shiripour.morteza@aalto.fi

Description: This project aims to develop an interactive multi-objective to graphical user interfaces.
Graphical user interface is a visual way for the user to interact with a computer via several graphical elements
(e.g. icons, menus, windows, and buttons) and its layout problem aims to find the best way to organize these
graphical elements on a fixed canvas. Organizing these layouts is challenging because there are a huge number
of different layout designs and the designers should also consider various limitations and criteria in their
layout design process. Hence, we intend to present a model and algorithm in order to automatically generate
layouts addressing different multiple tasks such as usability and aesthetic qualities. Moreover, we aim to guide
the algorithm to find solutions according to a decision maker preferences. [This work is associated with the
Finnish Center for AI (fcai.fi).]

Prerequisites: The successful intern should have a background in mathematical programming and
evolutionary multi-objective optimization, a strong previous background in coding in Python and machine
learning methods are beneficial (optional).

2108 - Inferring agent preferences from behaviour

Computer Science - Human-Computer Interaction, Machine

Field of study:
Learning, and Cognitive Science
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Andrew Howes andrew.howes@aalto.fi

P a g e 36 | 42
Description: The project will investigate reinforcement learning algorithms with which an observer can infer
the preferences of agents that forage for care-packages in a type of grid world that has varying terrain. Agents
differ in their preferences for different care-packages and in how costly they find different terrains to
traverse. Agents also differ in their cognitive capacities (e.g. their awareness of the care-packages and terrain).
The preferences and capacities of any particular agent are hidden from the observer but they do partly
determine the agent's behaviour. The task is for the observer to infer the agent's preferences given only the
agent's behaviour. [This work is associated with the Finnish Center for AI (fcai.fi).]

Paper: Jara-Ettinger, J., Schulz, L. E., & Tenenbaum, J. B. (2020). The naive utility calculus as a unified,
quantitative framework for action understanding. Cognitive Psychology.

Requirements: Cognitive modeling or reinforcement learning

2109 - Model-based eye tracking

Computer Science - Human-Computer Interaction, Machine

Field of study:
Learning, and Cognitive Science
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Xiuli Chen xiuli.chen@aalto.fi
Academic contact person:
Andrew Howes andrew.howes@aalto.fi

Description: Eye tracking is a technology that allows a computer to estimate where on a computer screen a
person is looking. It is used routinely in Human-Computer Interaction research, as well as in gaming and
virtual/augmented reality. The project will investigate whether the accuracy of eye tracking (how well it
determines what a person is looking at) can be improved with a model of how people use the eyes to gather
information. [This work is associated with the Finnish Center for AI (fcai.fi).]

Requirements: Cognitive modeling or reinforcement learning

2110 - Computational Graphical User Interface Evaluation

Field of study: Computer Science - Human-Computer Interaction (HCI)

School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Markku Laine markku.laine@aalto.fi

P a g e 37 | 42
Description: Graphical user interfaces (GUI) are omnipresent in our everyday life and play an important role
in how users interact with various electronic devices. Typically, evaluation of GUIs relies on personal
experience (subjective) and empirical testing (timeconsuming), and less so on computational modeling. To
address these concerns, we developed Aalto Interface Metrics (AIM), an online web service
(https://interfacemetrics.aalto.fi/) and codebase for computational evaluation of GUI designs.

The intern will contribute to the further development of AIM, including (i) the study and implementation of
new metrics and models and (ii) web development.
[This work is associated with the Finnish Center for AI (fcai.fi).]

Requirements: The intern is expected to have strong programming skills in Python and JavaScript, and
background in human-computer interaction.

P a g e 38 | 42
2111 - Combining optimization with Machine learning for UI design

Computer Science - Mathematics, Machine learning,

Field of study:
Optimization Techniques
School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Antti Oulasvirta antti.oulasvirta@aalto.fi
Academic contact person: Niraj Dayama niraj.dayama@aalto.fi

Description: This project aims to integrate Machine learning with Optimization techniques for solving
problems in User Interface design. Graphical user interface is the visual way for the user to interact with a
computer via several graphical elements (e.g. icons, menus, windows, and buttons) and its layout problem
aims to find the best way to organize these graphical elements on a fixed canvas. Organizing these layouts is
challenging because there are a huge number of different layout designs and the designers should also
consider various limitations and criteria in their layout design process. Hence, we intend to develop
techniques to handle UI layouts addressing different multiple tasks such as usability and aesthetic qualities.

Prerequisites:
The successful intern should have a background in mathematical programming, optimization techniques and
machine learning.

2112 - Interledger and Decentralised Identifiters for IoT Privacy

Field of study: Information technology

School: School of Electrical Engineering
Department: Department of Communications and Networking
Professor: Raimo Kantola raimo.kantola@aalto.fi
Academic contact person: Yki Kortesniemi yki.kortesniemi@aalto.fi

P a g e 39 | 42
Are you a developer interested in Internet-of-Things (IoT), distributed ledgers and interledger technologies,
and decentralised identifiers - in real-world applications?

We are now looking for summer interns to participate in the development work at Aalto University’s IoT-
NGIN project group in collaboration with experts around Europe. IoT-NGIN (https://iot-ngin.eu) is a large
European Union H2020 project that applies multiple novel technologies to develop more efficient, secure and
privacy preserving IoT solutions as part of the European Next Generation Internet.

Your work focuses on the development of privacy preserving solutions using interledger and decentralised
identifier technologies to support Digital Twins and other IoT applications. We except applicants to have:
• At least a bachelor’s degree in computer science or related field
• Good programming skills
• Solid experience with Python and Web service backend development
• Experience with distributed service design and API specification
• Good command of written and spoken English
• Ability to work effectively in an international environment

This position enables you to work on leading edge technologies with real-world applications in a fast-growing
field.

Department of Electronics and Nanoengineering

2201 - Organic Materials and Devices

Field of study: Physics, Materials Science, Chemistry

School: School of Electrical Engineering
Department: Department of Electronics and Nanoengineering
Professor: Caterina Soldano caterina.soldano@aalto.fi
Academic contact person: Caterina Soldano caterina.soldano@aalto.fi

The Organic Electronics Group is looking for a curious and talented student, either at BSc or MSc level, for the
Summer 2021, as part of the Aalto Science Institute internship program. You will contribute to the advance of
a novel class of organic devices (organic light emitting transistors); in particular, your role will be to develop
thin films, both dielectrics and organic materials, in order to study and improve device performances.
Experimental work will include growth of materials, surface characterization and device fabrication.
Our group is international, thus a good command of English is required.

Further information: Prof. Caterina Soldano (please visit our website for more info
https://organicelectronics.aalto.fi/)

P a g e 40 | 42
School of Chemical Engineering
Department of Chemistry and Material Science

3101 - Investigation Li-ion batteries electrode degradation processes

Field of study: Electrochemistry

School: School of Chemical Engineering
Department: Department of Chemistry and Material Science
Professor: Tanja Kallio tanja.kallio@aalto.fi
Academic contact person: Ekaterina Fedorovskaya ekaterina.fedorovskaya@aalto.fi

Description: In this project, investigation of ageing of Li-ion batteries comprising of advanced materials is
carried out using an accelerated aging test and related electrochemical and structural characterization
methods. The positive and negative electrode materials are extracted and investigated before and after ageing
using various structural and electrochemical characterization methods such as Fourier transform infrared
spectroscopy, Raman spectroscopy, X-ray photoelectron spectroscopy, Scanning electron microscopy,
Transmission electron microscopy, cyclic voltammetry, rate capability measurements, electrochemical
impedance spectroscopy. The work includes carrying above-mentioned experiments and/or interpreting the
obtained data.

Skills: Knowledge and experience on using of the above-mentioned characterization methods as well as
knowledge and experience on inorganic synthesis and electrochemical characterization methods are
appreciated. Furthermore, experience on using a glove box and assembling of electrochemical cells, such
supercapacitors and Li-ion batteries, is valued.

Education: The appropriate candidate has background in Chemistry, Material Science or related fields.
Completing courses on Analytical chemistry, Inorganic Chemistry, Organic Chemistry, Electrochemistry,
Physical Chemistry, Chemical Kinetics, Chemical Thermodynamics, Chemistry of Solids, Materials Science,
Materials for Electrochemical Engineering is beneficial.

P a g e 41 | 42
School of Engineering
Department of Civil Engineering

4101 - Computational analysis of microarchitectural solids and structures

Field of study: Computational mechanics

School: School of Engineering
Department: Department of Civil Engineering
Professor: Jarkko Niiranen jarkko.niiranen@aalto.fi
Academic contact person: Jarkko Niiranen jarkko.niiranen@aalto.fi

Description: The internship studies within the topic “Computational analysis of microarchitectural solids
and structures” will focus on the modelling of solids and structures -- such as beams, plates and shells --
having a microarchitecture, typically a periodic lattice or cellular microarchitecture. The peculiarities of the
thermomechanical behavior of these types of solids (e.g., the so-called size effects) can be captured via
(computational) homogenization and generalized continuum theories.

Requirements: The topic requires solid understanding of continuum mechanics and engineering
mathematics plus some experience in finite element analysis and Matlab-programming.

P a g e 42 | 42