Scribd
Upload
5 views

Assignment 1

Uploaded by

sakshamtyagi12a
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
5 views

Assignment 1

Uploaded by

sakshamtyagi12a
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

Soft Computing and Applications -

Assignment Questions
1. Explain the differences between 'soft' computing and 'hard' computing with relevant
examples.

2. Describe the architecture of a backpropagation neural network and its training algorithm.

3. How do fuzzy sets and fuzzy relations differ? Provide an example of each.

4. Explain the significance of fuzzification and defuzzification in fuzzy logic systems. Provide
examples of methods used in each process.

5. What is the role of membership functions in fuzzy logic? Describe at least two methods of
membership value assignment.

6. Outline the basic genetic algorithm (GA) framework and discuss the importance of
selection, crossover, and mutation operators.

7. Discuss the concept of neuro-fuzzy systems. How do they combine the strengths of both
neural networks and fuzzy logic?

8. Explain the working of an AI search algorithm, such as A* or DFS. How can these be
applied to solve optimization problems?

9. Compare and contrast the Mamdani and Sugeno types of fuzzy inference systems.

10. What are some real-world applications of soft computing techniques, particularly in the
field of artificial intelligence and machine learning?

11. Given a 3-layer neural network with 4 input nodes, 5 hidden nodes, and 2 output nodes,
calculate the total number of weights and biases in the network.

12. Consider a single-layer perceptron trained on the following data:

 Input: (1, 0), Desired Output: 1


 Input: (0, 1), Desired Output: 0
 Input: (1, 1), Desired Output: 1
 Input: (0, 0), Desired Output: 0
 Using a learning rate of 0.1, calculate the weight updates after one epoch.

13. A fuzzy control system generates the following fuzzy output set:

 Set A with membership values {0.2, 0.4, 0.6, 0.8}.


 Use the centroid method to calculate the crisp output if the universe of discourse
is {10, 20, 30, 40}.

14. A population in a genetic algorithm consists of 6 individuals with the following fitness
scores: {5, 10, 3, 8, 6, 2}. Using roulette wheel selection, calculate the probability of selecting
each individual.

15. Apply mutation on the binary string 110110 with a mutation probability of 0.1. How
likely is a single mutation, and if a mutation occurs, what would be the resulting string?

16.  A fuzzy inference system uses the following fuzzy rules:

 Rule 1: IF temperature is high (membership = 0.7) THEN fan speed is fast


(membership = 0.6)
 Rule 2: IF temperature is low (membership = 0.3) THEN fan speed is slow
(membership = 0.4)

 Using the max-min composition method, determine the inferred output fan speed.

17. Draw the functional block diagram of swarm intelligence system, explain the role of sub
blocks. Also represent the properties and application domain of swarm intelligence system.

18. Using McCulloch-Pitts neuron model, design a neural network for 2-input XOR functions.

19. Design a fuzzy logic-based decision-making system. Choose an application, such as a


temperature control system, washing machine settings, or a traffic signal controller.

 Define input variables (e.g., temperature, speed, traffic density).


 Define fuzzy sets and membership functions.
 Write fuzzy rules for decision-making.
 Implement the system using a programming language or software like Python

20. Develop a neural network that can classify a dataset, such as the MNIST handwritten
digits, flower species (Iris dataset), or a custom dataset.

 Use a machine learning framework (e.g., TensorFlow, PyTorch).


 Train and evaluate the network.
 Explore different network architectures (e.g., varying layers, activation functions).

You might also like