STUPIDSID
1 (a)
What is neural learning? Draw and explain the general neuron model.
8 M
1 (b)
Briefly discuss how neural networks are used for vector quantization and function approximation.
12 M
2 (a)
What is the approximate choice for the learning rate η in perceptron training algorithm?
5 M
2 (b)
What is the goal of the pocket algorithm? Explain with the help of the algorithm.
10 M
2 (c)
What is the termination criterion in perceptron training algorithm, if the given samples are not linearly separable?
5 M
3 (a)
What is back propagation? Explain the back propagation training algorithm with the help of a one-hidden layer feed forward network.
12 M
3 (b)
Explain the effect of momentum terms and number of samples used for training in back propagation algorithm.
8 M
4 (a)
Explain how the quickprop algorithm can be used to accelerate the learning process.
8 M
4 (b)
Briefly explain the network pruning algorithm.
4 M
4 (b)
Differentiate between the supervised and unsupervised methods of training.
8 M
5 (a)
Discuss briefly the two networks used for prediction problems.
12 M
5 (b)
Write the simple competitive learning algorithm for winner - take all networks and explain.
8 M
6 (a)
Explain the architecture of full counter propagation neural network with a neat diagram.
8 M
6 (b)
Explain how an unsupervised learning mechanism can be adopted to solve supervised learning tasks with the help of linear vector quantization (LVQ) algorithm.
12 M
7 (a)
What are Hopfield networks? Explain discrete Hopfield networks in detail.
6 M
7 (b)
What is simulated annealing? Briefly explain move - generation and move - acceptance.
8 M
7 (c)
Explain how bidirectional associative memory (BAM) can be used as hetero-associative memory.
6 M
8 (a)
Explain optimization using Hopfield networks for solving simultaneous linear equations.
8 M
8 (b)
Write the algorithm for generic evolutionary computation and explain termination criterion and initialization with respect to evolutionary algorithm.
12 M
More question papers from Artificial Neural Network