Scribd
Upload
16 views3 pages

Soft Computing Notes

Soft Computing encompasses techniques like fuzzy logic, neural networks, and genetic algorithms that handle imprecision and uncertainty, contrasting with Hard Computing's reliance on precise solutions. The document outlines key concepts such as the differences between Artificial Neural Networks and Biological Neural Networks, various activation functions, and the architecture of ANNs. Additionally, it covers training algorithms, fuzzy set operations, and the processes of fuzzification and defuzzification.

Uploaded by

kidslove229
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
16 views3 pages

Soft Computing Notes

Soft Computing encompasses techniques like fuzzy logic, neural networks, and genetic algorithms that handle imprecision and uncertainty, contrasting with Hard Computing's reliance on precise solutions. The document outlines key concepts such as the differences between Artificial Neural Networks and Biological Neural Networks, various activation functions, and the architecture of ANNs. Additionally, it covers training algorithms, fuzzy set operations, and the processes of fuzzification and defuzzification.

Uploaded by

kidslove229
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
You are on page 1/ 3

Soft Computing Notes

1. What is Soft Computing & Differences with Hard Computing

Soft Computing refers to computing techniques that are tolerant of imprecision, uncertainty, partial

truth, and approximation. It includes fuzzy logic, neural networks, and genetic algorithms. Hard

Computing relies on precise, binary logic, and exact solutions.

Figure: Soft vs. Hard Computing (Include Venn diagram or comparison table).

2. Difference between ANN & BNN (Biological Neural Networks)

ANN (Artificial Neural Network): Artificially designed systems inspired by biological neurons. BNN

(Biological Neural Network): Real neural networks in animals or humans.

Differences: ANN is mathematical; BNN is biological; ANN has fixed architecture; BNN is highly

plastic.

3. Explain any 4 Activation Functions with Examples

Linear: f(x)=x

Bipolar: f(x)=1 if x>0, -1 if x<=0

Binary: f(x)=1 if x>threshold, else 0

Sigmoid: f(x)=1/(1+e^-x)

Include small plots or curves for each function.

4. Architecture of ANN

Basic architecture: Input layer, hidden layer(s), output layer.

Figure: Diagram showing nodes, connections, weights.

5. Difference: Supervised vs. Unsupervised Neural Network


Supervised: Uses labeled data, learns by example.

Unsupervised: Uses unlabeled data, finds patterns or clusters.

6. ADA LINE Training Algorithm (with Numerical)

ADA LINE: Adaptive Linear Neuron; updates weights using error correction.

Numerical example: Given input and weights, compute output, error, adjust weights.

7. Perceptron Learning Model (Training Algorithm)

Steps: Initialize weights, for each input compute output, update weights if error, repeat until

convergence.

8. Backpropagation Algorithm (Steps Only)

1. Forward pass: compute outputs.

2. Compute error.

3. Backward pass: propagate error.

4. Update weights using gradient descent.

9. Hetero-associative Memory Network (Numerical)

Maps input patterns to different output patterns.

Numerical: Given X and Y pairs, compute weight matrix, recall output.

10. Difference between Crisp Set & Fuzzy Set

Crisp Set: Binary membership (0 or 1).

Fuzzy Set: Degrees of membership (0 to 1).

11. Fuzzy Set Operations + Numerical

Operations: Union, Intersection, Complement.

Numerical: Example with fuzzy membership values.


12. Properties of Fuzzy Relations

Reflexivity, symmetry, transitivity, max-min composition.

13. Fuzzy Membership Function & Properties of Fuzzy Set

Membership function: Defines degree of membership.

Properties: Normality, convexity, support, core.

Figure: Triangular or trapezoidal membership functions.

14. Fuzzification + Example

Fuzzification: Converting crisp input into fuzzy values.

Example: Weight of people, define linguistic variables (light, medium, heavy), plot membership

functions.

15. Defuzzification Definition & Method

Defuzzification: Converting fuzzy results into crisp output.

Methods: Centroid, bisector, mean of maxima, etc.

You might also like