JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY HYDERABAD B.Tech III ARTIFICIAL NEURAL NETWORKS Previous Question Paper (August/September - 2024 - R16)

Define a neural network.

What is the human brain's role in the study of neural networks?

What are learning curves in neural networks?

How does a perceptron relate to a Bayes classifier in a Gaussian environment?

Define the term “unsupervised learning”.

List the effects of cross validation on model performance.

What is the main purpose of a Self-Organizing Map (SOM)?

List the key properties of feature maps in SOMs.

Define a dynamical system.

How are Hopfield models utilized in neuro-dynamical research?

Explain in detail the concept of a neural network and its components.

Illustrate the role of knowledge representation in neural networks.

Summarize the concept of competitive learning.

Explain the process of error correction learning and its significance.

Describe the XOR problem and explain why it cannot be solved by a single layer perceptron.

Illustrate the working of linear least square filters with an example.

Discuss various heuristics used in the training of multilayer perceptrons and their impact on learning.

Evaluate the role of learning curves in model training and performance evaluation.

Supervised Learning

Hessian Matrix.

Demonstrate the accelerated convergence in back propagation.

Classify different network pruning techniques used in neural networks.

Determine the neighborhood function used in SOM training.

Outline the process of training a Self-Organizing Map (SOM) for image recognition tasks.

Categorize the various applications of SOMs.

Compare the Self-Organizing Map (SOM) with Learning Vector Quantization (LVQ).