0% found this document useful (0 votes)
13 views5 pages

Soft Computing

The document contains a question bank focused on soft computing models, including multiple-choice questions and short/long answer type questions related to artificial neural networks (ANN), fuzzy systems, genetic algorithms (GA), and various learning mechanisms. It covers topics such as back-propagation networks, activation functions, and the differences between fuzzy and crisp sets. Additionally, it includes design tasks for neural networks and fuzzy inference systems, as well as explanations of key concepts in soft computing.

Uploaded by

habuldas23
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
0% found this document useful (0 votes)
13 views5 pages

Soft Computing

The document contains a question bank focused on soft computing models, including multiple-choice questions and short/long answer type questions related to artificial neural networks (ANN), fuzzy systems, genetic algorithms (GA), and various learning mechanisms. It covers topics such as back-propagation networks, activation functions, and the differences between fuzzy and crisp sets. Additionally, it includes design tasks for neural networks and fuzzy inference systems, as well as explanations of key concepts in soft computing.

Uploaded by

habuldas23
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

SOFT COMPUTING MODEL QUESTION

BANK
MCQ

1 Back-Propagation Network is based on

Hebbian learning rule (b) Perceptron learning rule,


(c) Delta learning rule (d) None of these

2 ANN is generally used for

(a) optimizing (b) searching (c) learning and adaptation (d) decision making

3 Activation function is used in

Fuzzy systems (b) ANN systems (c) GA systems (d) None of these

4 The gradient descent rule mostly is used in

(a) Mc-Culloch Pitts learning (b) Hebb learning


(c) Back propagation learning (d) ADALINE learning

5 Error correction learning is

a) Unsupervised b) Supervised
c) Reinforcement d) Both (a) & (b)

Mc-Culloch Pitts Model uses

Tan hyperbolic function b) signum function


c) step function d) sigmoid function

Radial basis function uses

(a) signum function b) step function


(c) sigmoid function d) Gaussian function

MADALINE stands for

Many Adaptive Linear Neuron b) Many Additive Linear Neuron


c) Many Associative Linear Neuron d) Many Adaptive Derivative Linear Neuron

A neural network has weights 2 and 4 having inputs 5 and 10. The transfer function is linear
with constant of proportionality is 5. The output is

a) 250 b) 50 c) 2500 d) 25

Single layer Perceptron is used for

error minimization (b) back propagation


linear separability (d) annealing

Which of the following is not suitable for back propagation nets?


a) sigmoid b) hyperbolic tangent c) step function d) None of the above

‘winner-takes-all’ strategy is used in is

Competitive learning b) Hebbian learning


c) Boltzman learning d) Memory based learning

Functional values of Bipolar sigmoid function is

0 to 1 (b) -1 to 1 (c) any positive values (d) - ∞ to + ∞

The competitive rule is suited for

(a) Reinforced neural network (b) Supervised neural network


(c) Unsupervised neural network (d) none of these

A Perceptron is

A single layer feed-forward neural net with pre-processing


an auto association neural net
a double layer auto association neural net
(d) none of these

Bias is

Weight on a connection from a unit having activation 1


Weight on a network having activation 2
Weight on a function having activation 1
Weight on a function having activation 0
When the fitness values of Chromosomes in GA differ very much the best reproduction
operator is

(a) Roulette wheel selection (b) Tournament selection,


(c) Rank selection, (d)None of these

A fuzzy data with membership value 1.0 is called

(a) crossover point (b) core (c) centre (d) Both (b) and (c)

A fuzzy data with membership value 0.5 is called

(a) crossover point (b) core (c) centre (d) support

GA is used for

(a) optimization (b) searching (c) adaptation (d) Both (a) and (b)

Which of the following belong to Hybrid soft computing

(a) Fuzzy systems (b) ANN systems (c) GA systems (d) Neuro-Fuzzy system

SHORT/LONG ANSWER TYPE QUESTIONS

Define soft computing. How it is related to AI and machine learning? Name different soft
computing techniques. State their Applications.

Name different learning mechanisms used in ANN.

State the role of weight, bias and activation function in neural network design with example.
Design a McCulloch-Pitts Neural net to realize the XOR function.

Design a Hebb net to implement logical OR function with bipolar inputs and targets. Define
Perceptron learning rule.

Explain the difference between single layer Perceptron and multi layer Perceptron with
diagram.

Draw a 5,3,4,2 multilayer feed forward neural network.

The Exclusive –OR function is not linearly separable. Explain with diagram. Design a multi
layer Perceptron network to solve XOR problem
Implement AND function with inputs and bipolar targets using Perceptron training
algorithm upto 2 epochs.

Explain Back propagation algorithm with suitable diagram. Explain learning factors of Back
propagation network.

Find suitable weight and bias values of a two input Radial basis function neural network to
solve XOR problem.

Implement AND/OR/EX-OR function using Hebb Network (Use bipolar inputs and targets)
State some applications of neural network.

State difference between Fuzzy set and Crisp set. Explain Fuzzy membership function with
an example.

What is the difference between fuzziness and randomness? Explain Fuzzy Inference System
(FIS) with an example.

Develop an FIS (Mamdani) model for controlling temperature in an air conditioner. State
different Defuzzification methods and explain any one of them.

Draw a flowchart for Genetic Algorithm and explain it.

Explain Encoding / Crossover operator used in GA.

Explain the following terms wrt GA:

population, fitness function, Elitism,Chromosomes

Explain Fuzzy membership function with an example to implement “LOW”, “MEDIUM” and
“HIGH”.

With an example explain the approach of fuzzy rule base to address the linguistic inputs to a
system.

Short Notes:

Activation function Correction learning, Hebbian learning, Competitive learning, Gradient


descent learning, Regression,

McCulloch-Pitts model,
Perceptron, Multi Layer Perceptron,

Madaline networks

Fuzzy membership functions,

Fuzzy Rule based Systems, Fuzzy Inference System (FIS), Defuzzification

Roulette wheel selection / Tournament selection,

Ant Colony Optimization (ACO) / Particle Swarm Optimization (PSO).

You might also like