Course BUAN/MIS/OPRE 6341

Course Title Applied Machine Learning

Instructor Uri Smashnov
Term Spring 2024
Meetings Lecture with Virtual Attendance (Online)

Professor Contact Information

Instructor: Uri Smashnov
Email Address uri.smashnov@utdallas.edu
Office hours: virtual office hours via MS Booking: TBD

Teaching Assistant: TBD

Email:
Office hours:

Course Modality and Expectations

Instructional Mode Lecture - recorded
Lectures will be pre-recorded and class materials made
available on a Module by module basis.
Lecture Time
Projects will be introduced in the synchronous session that will
be recorded and posted on eLearning.
Course Platform Course materials will be posted the eLearning page of the
course.
Asynchronous Learning Posted on on eLearning
Guidelines

Course Pre-requisites, Co-requisites, and/or Other Restrictions

Please consult CourseBook and your academic advisor as of August 5th, 2023 the
pre-requisites are listed as following: Prerequisites: (BUAN 6356 or BUAN 6383 or
MIS 6386 or BUAN 6324 or MIS 6324 or OPRE 6399) and (OPRE 6359 or BUAN
6359).

The course is taught in Python. Students are expected to have prior programming
and Python experience.

BUAN/MIS/OPRE 6341 Course Syllabus Page 1

Course Description

In this advanced course, we will delve into various machine learning models tailored for
business analytics. Our focus will encompass both supervised and unsupervised models,
intricate non-linear regression techniques, resampling methods, and introduction to
neural networks and AI-driven models essential for data analytics. Python, a leading
programming language in this field, will be our primary tool. Given the course's
advanced level, a proficient understanding of Python is assumed, and we will allocate
minimal time to revisiting Python basics. Additionally, we will explore crucial aspects of
machine learning interpretability, including global and local Explainability, model
debugging, and fostering trust in model outputs.

Student Learning Objectives/Outcomes

Upon completion of this course, students will have the skills to:
1. Distinguish between supervised learning methods (such as classification) and
unsupervised learning methods (like clustering).
2. Select the appropriate machine learning technique for a given dataset and
business scenario.
3. Skillfully perform feature engineering to optimize machine learning models and
proficiently use Python for analytical purposes.
4. Design and execute machine learning experiments aimed at minimizing
overfitting and achieving the best possible results with the available data in
relation to specific business challenges.
5. Grasp and implement concepts related to the interpretability of machine
learning models.

Required Textbooks and Materials

Required Texts
Textbooks are not required for this course. Supporting material will be provided through
eLearning course page.

Required Hardware and Software

The lecture materials are tested only with Windows 10 and latest MacOS running Intel
Chip. Mac with M1/M2 chip is not tested. Some libraries might require compilation and
additional steps to be installed on Mac OS M1/M2 and students should contact
professor or TA if help with installation is needed.
Required Memory: at least 8 GB memory is required, 16 GB memory is recommended.
Students can rent/provision Linux/Windows box on one of the Cloud providers and use
for the class if personal computer doesn’t meet memory requirements.

BUAN/MIS/OPRE 6341 Course Syllabus Page 2

Suggested approach: provision AWS EC2 Ubuntu box (t3a.xlarge) with Spot pricing and
only “start” it while using it, rest of the time only storage charges will apply. It is possible
to open tunnel to Jupyter notebook only from your PC to make the setup very secure.
Neural nets module will be taught using Google Colab. Premium ($10 per month – only
1 month will be needed) subscription is advised to enable GPU use.

Software: Python 3.10.11 - Additional packages you need to have for this course are
H2O-3, numpy, scipy, sklearn, pandas, matplotlib, seaborn, graphviz, TensorFlow, and
keras. Exact version of Python and the libraries will be posted on eLearning.
H2O-3 has Java version 1.8 or later requirement.
Use of Python virtual environment is mandatory – details are provided in the in
eLearning.

Recommended Texts

Introduction to Machine Learning with Python

ISBN 978-1-4493-6941-5
Publisher O'Reilly Media, Incorporated
https://learning.oreilly.com/library/view/introduction-to-
machine/9781449369880/

Introduction to Machine Learning

Author Ethem Alpaydin and Francis Bach
ISBN 9780262043793
Publisher MIT Press

Machine Learning at Scale with H2O

Authors: Gregory Keys and David Whiting
ISBN: 978-1800566019
Online Book (Free with UTD login):
https://learning.oreilly.com/library/view/machine-learning-at/9781800566019/

Online Sklearn and H2O-3 documentation and tutorials.

Textbooks and some other bookstore materials can be ordered online or purchased at
the UT Dallas Bookstore.
Some books are available online and are free with UTD login. Playlist can be found here:
Oreilly Learning Books playlist

BUAN/MIS/OPRE 6341 Course Syllabus Page 3

Technical Requirements
In addition to a confident level of computer and Internet literacy, certain minimum
technical requirements must be met to enable a successful learning experience. Please
review the important technical requirements on the on the Getting Started with eLearning
webpage.

Course Access and Navigation

The course can be accessed using your UT Dallas NetID account on the eLearning
website. Please see the course access and navigation section of the Getting Started with
eLearning webpage for more information.

To become familiar with the eLearning tool, please see Student eLearning Tutorials
webpage.

UT Dallas provides eLearning technical support 24 hours a day/7 days a week. The
eLearning Support Center includes services include a toll-free telephone number for
immediate assistance (1-866-588-3192), email request service, and an online chat
service.

Communication
All class discussions will be performed on MS Teams under appropriate channel. For
example, we will use Project channel for Project discussions. Use of email is appropriate
for personal questions only.

Distance Learning Student Resources

Online students have access to resources including the McDermott Library, Academic
Advising, The Office of Student Accessibility, and many others. Please see the eLearning
Current Students webpage for more information.

Server Unavailability or Other Technical Difficulties

The University is committed to providing a reliable learning management system to all
users. However, in the event of any unexpected server outage or any unusual technical
difficulty which prevents students from completing a time sensitive assessment activity,
the instructor will provide an appropriate accommodation based on the situation.
Students should immediately report any problems to the instructor and also contact the
eLearning Help Desk. The instructor and the eLearning Help Desk will work with the
student to resolve any issues at the earliest possible time.

BUAN/MIS/OPRE 6341 Course Syllabus Page 4

Student Assessments

Grading Criteria:
In line with the applied nature of this class, a large portion of the assessment will be made
through homework. There will be 2 project assignments. The homework will contain
some theory questions, but the majority of the material will involve implementing the
different methods that we cover in class using the computer package. There will be one
take-home exam. The breakdown will be:

Item Point
Project 1 15
Project 2 25
Weekly Labs 30
Final Quiz 30
Total 100

Grading Scale:
Relative grading is a system of assessment to determine students’ grades by comparing
them against those of their peers. Unlike the system of absolute grading, where a
student’s score on a given test or assignment directly converts into a letter grade (for
example, 93/100 is A, etc.), relative grading means students’ marks fluctuate depending
on how they did compared to others in class. Relative grading refers to a system of
evaluation that allows educators to convert the outcomes of a student’s test, project or
assignment and adjust that final grade in relation to grades from other students in the
course. Relative grading is similar to bell curving or grading on a curve, and considers
the highest score as the baseline (A), relatively adjusting all others compared to that
score. Student should earn a passing grade for each project and exam grading components
in order to be considered for a letter grade in the range of C to A.

Note: this grading system is following the UTD/JSOM policy to keep the class grade
average between B to A-.

Grade Range of marks

A 80th percentile and above. A grade of 91 or above will guarantee an A.
A- 60th to 80th percentile.
B+ 40th to 60th percentile.
B 20th to 40th percentile.
B-/C Below 20th percentile.
F Failing grade in any grading component. A grade of 60 or below out of 100
is a failing grade.

BUAN/MIS/OPRE 6341 Course Syllabus Page 5

Accessing Grades:
Students can check their grades by clicking “My Grades” on the course menu after the
grade for each assessment task is released.

Projects:
There are 2 individual projects for this course. Project details will be posted in the
course.

Labs:
There are Labs as individual assessment activities.

Quiz:
There will be one Quiz at the end for this course.
• The quiz will be held during regular lecture time on the last week of the class.
• We will be using the Honorlock system to monitor the quiz.
• Details of the quiz and the grading rubric will be posted on eLearning.

BUAN/MIS/OPRE 6341 Course Syllabus Page 6

Course Outline/Academic Calendar

Week of Topic/Lectures Details/Materials Assignment/Activity

Week 1 Course access and self- Syllabus Lab 0 – Python
1/15/2024 introduction environment set-up
Module 1: Introduction to Scikit-learn: Machine Learning in Lab 1
Pandas, Python Python
environment set-up
Week 2 Module 2: Introduction to H2O-3 – documentation Lab 2
1/22/2024 Machine Learning

Week 3 Module 3: Supervised • Supervised Learning – types and Lab 3

1/29/2024 Learning approaches
• Decision trees
Week 4 Module 3 continued Linear models Project 1 – Posted
2/5/2024 Lab 4
Week 5 Module 3 continued Intro to Ensemble Learning. Random Lab 5
2/12/2024 Forest
Week 6 Module 4: Model Tuning, Lab 6
2/19/2024 Evaluation and Selection
Week 7 Module 4 continued
2/26/2024
Week 8 Module 5: Ensemble Combining Multiple Learners
10/11/2023 Learning
Week 9 Module 5 continued
3/4/2024
Spring Break
Week 10 Module 6: ML Project 2 - Posted
3/18/2024 Interpretability
Week 11
3/25/2024 Module 6: continued Lab 7
Week 12 Module 7 – model
4/1/2024 debugging
Week 13 Module 8: Dimensionality Dimensionality Reduction: PCA and
4/8/2024 Reduction GLRM (H2O) Lab 8
Week 14 Module 9: Neural • Perceptron Learning
4/15/2024 Networks • Weighted Networks - The perceptron
• The Perceptron: A probabilistic model
for information storage and organization
in brain Lab 9
Week 15 Overview of LLMs Overview of Large Language Models
4/22/2024 and how to use LLM for
Classification/Regression

BUAN/MIS/OPRE 6341 Course Syllabus Page 7

Course Policies
Make-up exams: No makeup exam is allowed in this course.
Extra Credit: There is not any extra credit activity for this course.

Late work policy

• All submissions, including labs and projects, will have a grace period of 10 hours.
• Submissions after the grace period will not be accepted.
• The grace period is intended to help students who experience technical problems
with eLearning.
• The grace period is not intended to give students extra time to work on
assignments.
• Plan to submit your work early if you know you will not around the due date.
o Submission links will be available at least 1 week in advance of the due
date.
o This gives you plenty of time to submit your work early, even if you have
personal commitments, work load, or UTD work load that will make it
difficult to complete the assignment on time.
Class Participation and Camera use

• Optional: Class participation during Project introduction. The sessions will be

recorded and conducted via MS Teams. Exact date/time and links to join will be
posted in the eLearning
• Students are encouraged to participate in the discussion on class MS Teams
channel.

Comet Creed

This creed was voted on by the UT Dallas student body in 2014. It is a standard that Comets choose
to live by and encourage others to do the same:

“As a Comet, I pledge honesty, integrity, and service in all that I do.”

UT Dallas Syllabus Policies and Procedures

The information contained in the following link constitutes the University’s policies and
procedures segment of the course syllabus. Please visit UT Dallas Syllabus Policies webpage for
these policies.

The descriptions and timelines contained in this syllabus are subject to

change at the discretion of the Professor.

BUAN/MIS/OPRE 6341 Course Syllabus Page 8