Cochin University of Science and Techology (CUST) 2005-7th Sem B.Tech Information Technology ,, IT 701 Neural Computing - Question Paper

Wnc:

3 Hour*

Maximum Marks: 100

1

>

Explain the characteristic* of artificial Dcura] network*.

(9)

b)

Describe the Me Cullocfc-Pttu model of a neuron.

(6)

e)

Explain the Hebbian Leamiag.

OR

(5)

II

>

Distinguish between supervised aad uasupervised learn mg

(10)

b)

Explain linearly inseparable problem with an example.

(10)

III

)

Describe the different activation functions wtucb function is used in back propagation training algorithm. Why?

(10)

b)

Explain the pcrceptron training algorithm.

OR

(10)

IV

)

Explain how the weight* are adjusted in the output layer of a neural network using back propagation training algorithm.

(10)

b)

Deccnbe the ADALINE model.

(10)

V

)

Explain the characteristics of ART.

(10)

b)

Describe the initializaikm and training in ART operation.

OR

(10)

VI

*)

Describe the major phases in ART classification process.

(10)

b>

Explaio the functional modules of ART network configuration.

(10)

VII

)

Explain with diagram recumn: and non recurrent networks

(10)

b)

Describe how data is stored and retrieved in BAM.

OR

(10)

VIII

)

Describe the stability condition in a Hopfield network.

(10)

b)

Explain any ooe application for the Hop field network.

(10)

IX

)

b)

Describe Kobooen layer and Grotsberg layer in a feed forward counter propagation network.

Explain what happens when the initial weight vetton are randomly chosen in a Kobooen self organizing map.

OR

(10)

(10)

X

>

Describe neocognitron model.

(10)

b)

Explain ibe role of excitatory neuron and inhibitory neuron in the cognition model

(10)