Scribd
Upload
3 views2 pages

Nural Network

This document outlines the examination structure for the Neural Networks course at MLR Institute of Technology, including details about the question paper format, which consists of two parts: compulsory questions in Part A and unit-based questions in Part B. It specifies the marks distribution and includes various topics related to neural networks, such as backpropagation, memory mechanisms, and the significance of different algorithms. The examination is scheduled for November 2023 and is designed to assess students' understanding of key concepts in neural networks.

Uploaded by

dhanushkuna333
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views2 pages

Nural Network

This document outlines the examination structure for the Neural Networks course at MLR Institute of Technology, including details about the question paper format, which consists of two parts: compulsory questions in Part A and unit-based questions in Part B. It specifies the marks distribution and includes various topics related to neural networks, such as backpropagation, memory mechanisms, and the significance of different algorithms. The examination is scheduled for November 2023 and is designed to assess students' understanding of key concepts in neural networks.

Uploaded by

dhanushkuna333
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

H.T NO.

R 2 MLR20
Course Code: A5AI21
MLR INSTITUTE OF TECHNOLOGY
(Autonomous)
IV B.Tech. I Semester Regular Examination November-2023
NEURAL NETWORKS
(CSM)
Time: 3 Hours. Max. Marks: 70
Note: 1. This question paper contains two parts A and B.
2. Part -A is Compulsory which carries 20 marks. Answer all Questions in part A.
3. Part -B consists 5units.Answer any one question from each unit. Each question
carries 10 marks and may have a, b, c as sub questions.
PART- A 10 x 2M = 20Marks
a) Define neural network and its functions. 2M CO1 BL1
b) Analyse the significance of the Boltzmann learning approach. 2M CO1 BL4
c) State backpropagation algorithm. 2M CO2 BL1
d) Define learning curves, and how do they reflect the performance 2M CO2 BL2
of single-layer perceptron in different scenarios.
e) List the key advantages of using the backpropagation algorithm 2M CO3 BL1
for training neural networks.
1 f) List the different cross-validation techniques commonly 2M CO3 BL1
employed in practice.
g) Discuss the properties of the feature map generated by the Self- 2M CO4 BL2
Organization Map (SOM) algorithm.
h) State Self-Organization Map (SOM) in the context of 2M CO4 BL1
unsupervised learning and pattern recognition.
i) Define dynamical systems. 2M CO5 BL1
Discuss the strategies for manipulating attractors as a recurrent
j) network paradigm in the context of neurodynamics and 2M CO5 BL2
computational neuroscience.
PART- B 5 x 10M = 50Marks
Discuss the various elements or components of a biological
a) neuron, and how are they represented in artificial neural 5M CO1 BL2
2 networks.
b) Analyse how neural networks incorporate memory mechanisms 5M CO1 BL4
to retain and recall learned information over extended periods.
OR
3 a) Illustrate process of representing neural networks as directed 5M CO1 BL3
graphs, and also specify the significance of this representation.
Determine the significance of the Boltzmann learning approach
b) in enhancing the learning capacity and generalization of neural 5M CO1 BL3
networks.
Page 1 of 2
4 a) Explain the various unconstrained organization techniques 5M CO2 BL2
employed in the context of single-layer perceptron.
Outline how output representation and decision rule determined
b) in a multilayer perceptron, and how does it affect the model's 5M CO2 BL4
predictive capabilities.
OR
a) Describe the concept of linear least square filters and their 5M
significance in the context of single-layer perceptron. CO2 BL2
5 Explain the heuristics involved in training a multilayer perceptron
b) and assess their impact on the model's performance and 5M CO2 BL4
convergence.
Outline the key steps involved in the backpropagation process,
6 and how does it contribute to the efficient training of deep 10M CO3 BL4
learning models.
OR
a) Discuss
their
the various techniques used for network pruning and 5M CO3 BL2
impact on the model's efficiency and performance.
7
b) Describe the concept of accelerated convergence in the context 5M CO3 BL2
of neural network training.
Discuss the concept of adaptive pattern classification and its
8 significance in the context of machine learning and neural 10M CO4 BL2
networks.
OR
Illustrate how computer simulations used to assess the
a) robustness and scalability of feature mapping models under 5M CO4 BL4
9 different input conditions
List the fundamental characteristics and key differences
b) between the two basic feature mapping models used in machine 5M CO4 BL2
learning

a) Analyse how stability of equilibrium states influence the 5M CO5 BL4


10 robustness and reliability of neural network behaviour
b) Explain the Hopfield model and its significance in understanding 5M CO5 BL2
associative memory and pattern completion in neural networks.
OR
Illustrate different types of attractors’ shape the dynamics of
11 neural networks, and Specify their implications for memory 10M CO5 BL3
retrieval and pattern recognition. Provide examples to illustrate
the concept of attractors in the context of neurodynamics.
---oo0oo---
Page 2 of 2

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy