Soft Computing Solved Question Paper – To Score Better – RGPV-CS-801

CS-801 (GS)

B.E. VIII Semester Examination, June 2020

Grading System (GS)

Soft Computing

Time : Three Hours

RGPV-CS-801 – Soft Computing – Solved Question Paper
1. Define the following:Description Links
i) Artificial IntelligenceLink 1
What is artificial intelligence?

– Artificial intelligence (AI) refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions.

What are some real life examples of artificial intelligence application?

Self-driving car project of UBER,

Google’s search algorithm that gives exact results for different queries,

Cancer detection application that detects cancer based in images,

Image detection e.g. google lens, Amazon Rekognition etc. to extract information out of images.

– Text mining: Finding useful information from text data etc.

– Language translation e.g. google translate

– Audio to text converter
ii) Soft ComputingLink 1
Basis of soft computing?

The concept of soft computing is based on learning from experimental data.

It means that soft computing does not require any mathematical model to solve the problem.

And it provides approximate results to complex world problems.

In string matching say of ‘ABC’ and ‘ABCD’ traditional algorithms will give answer in Yes or No, whereas using soft computing we can tell the % of matching between the two.

What are some applications of soft computing?

Handwriting recognition as this is also an approximate match considering that every person has it’s own unique writing style.

– In gaming product like Poker to find the winning cards etc.
b) Write and explain AO* algorithm.Link 1
When a problem can be divided into a set of sub problems, where each sub problem can be solved separately and a combination of these will be a solution, AND-OR graphs or AND – OR trees are used for representing the solution. 

i.e. using combination of two solutions to solve one problem.
2. What are the difference in learning approach of counter propagations network to fead forward network.Link 1
3. Write a short note on:
i) Associative memoryLink 1
What is associative memory?
– It is a special type of memory that is optimised for performing searches through data, as opposed to providing a simple direct access to the data based on the address.

What are advantages of associative memory?

It requires more RAM, but has several advantages like

– Speeds up the search in database

– Good for parallel searches etc.
ii) Recurrent networkLink 1
What is main characteristic of recurrent network?

It is the first algorithm that remembers it’s input due to an internal memory, which makes it perfectly suited for machine learning problems that involve sequential data

What is the benefit of using recurrent neural network (RNN)?

The benefit of RNN is that while making a decision it considers current input as well as what it has learned from previous inputs.

So, we get enhanced algorithm as it is using learning from previous inputs as well.

What are some real life application examples of recurrent neural network (RNN)?

Some examples:

– Apple’s siri

– Google’s voice search

– Time-series analysis etc.
4. Define the architecture of a perceptron? What do you mean by linear separability?Link 1
What is a perceptron?

A perceptron is a neural network unit (an artificial neuron) that does certain computations to detect features or business intelligence in the input data.

By business intelligence here it means finding valuable insights from the data that is otherwise not easily identifiable just by seeing into the data.

A Perceptron is an algorithm for supervised learning of binary classifiers, where supervised learning is where input as well as it’s output is provided to an algorithm to learn from it and later on applying the learning from it to unseen data and binary classifiers are 1,0 type classifiers i.e. with only two options.

Perceptron Learning Rule states that the algorithm would automatically learn the optimal weight coefficients.

The input features are then multiplied with these weights to determine if a neuron fires or not.

Neural networks works in this way where between input and output multiple layers are present each carrying different weights, then inputs are multiplied by those weights to reach to final output.

Linear separability is something that is linearly separable i.e. either by point, line, 2d plane etc.and there are two classes lying on each side.

6. What is soft computing? Compare soft computing with hard computing.
Explain various soft computing techniques.
Link 1
Link 2
For soft computing refer second part of 1st question.

Refer the links for difference between soft and hard computing, it is very well explained there.
7. What do you mean by Production system? What are various types of production system? Write down its characteristics.Link 1
What comprises production system?
Production systems are machinery used in systematic way to produce goods or services in an industry.

In this resources are transformed to final goods or services.

These methods have evolved very much over the last century, earlier it used to be manual and now have been automated for e.g. using automated panels that shows machine readings, information at real time basis engineers operates it from a closed environment instead of operating it from field.

Also, this automated systems have increased productivity and non-stop production 24×7 round the year with minimal outage.

What are several types of production systems?

Several of production systems are:

– Batch: One set of quantity is produced and then another is put into system for transformation

– Continuous: Works non-stop supply of raw material and final output remains on without interruption

– Project: Is a one-shot system to produce one of a kind product, so reproducibility of same product is not easy

8. Write short notes:(Any two)
a) CrossoverLink 1
Link 2
What is crossover?
It’s a term originated from biology, where chromosomes of two individuals are crossover or combines to give life to new improved chromosome offspring.

What are several types of crossover?
There are several types of crossover for e.g. (refer the second link for visualisation os the below crossover methods)

– Single point crossover: One random point is selected and the part to the right of that point is swapped between two chromosomes

– Two point crossover: Two random points are chosen on the individual chromosomes (strings) and the genetic material is exchanged at these points.

– Uniform crossover: In uniform crossover, typically, each bit is chosen from either parent with equal probability. Other mixing ratios are sometimes used, resulting in offspring which inherit more genetic information from one parent than the other.
b) Mutation operatorLink 1
Link 2
What is mutation?
In simple terms, mutation may be defined as a small random tweak in the chromosome, to get a new solution.

It is used to maintain and introduce diversity in the genetic population.

For different types of operator refer the first link, which describes 5 type of mutation operators i.e.
– Bit flip
– Random setting
– Swap
– Scramble
– Inversion mutation

Each has it’s own method of creating change in chromosome.
c) Reproduction phase of genetic algorithmLink 1
Basically, reproduction of offspring happens in two ways
a. Crossover
b. Mutation

Both are described in detail in point (a) and (b) above, so please refer that.
d) Fuzzy membership functionLink 1

Other relevant articles:

13 thoughts on “Soft Computing Solved Question Paper – To Score Better – RGPV-CS-801”

Leave a Comment