Scribd
Upload
33 views

Course Code::: Neural Network and Fuzzy Logics ECE3008

This document contains information about the course "Neural Network and Fuzzy Logics" with course code ECE3008. It includes two questions related to neural networks and ART-1 clustering. The first question involves using a multilayer perceptron neural network to approximate a function, and the second question implements ART-1 learning to cluster six patterns into three clusters with different vigilance parameter values. Code is written in MATLAB to solve both questions and results are plotted or displayed.

Uploaded by

sankeerth
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
33 views

Course Code::: Neural Network and Fuzzy Logics ECE3008

This document contains information about the course "Neural Network and Fuzzy Logics" with course code ECE3008. It includes two questions related to neural networks and ART-1 clustering. The first question involves using a multilayer perceptron neural network to approximate a function, and the second question implements ART-1 learning to cluster six patterns into three clusters with different vigilance parameter values. Code is written in MATLAB to solve both questions and results are plotted or displayed.

Uploaded by

sankeerth
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/ 11

​ Course Name​: Neural Network and Fuzzy Logics

​ Course Code: ​ ECE3008

Prof.Neha Gupta -

Name​: M.Sankeerth
Reg No:​ 17BEV7002
1. Use back propagation algorithm to approximate the following function ()=1+cos
(pi/4) −5≤≤5. The training stops when the Mean Square Error (MSE) reaches zero
or a predefined maximum number of epochs is reached. Design Multilayer Perceptron
neural network with 1 hidden layer and 3 neurons in hidden layer. Use log-sigmoid and
tangent sigmoid transfer function in hidden layer and output layer respectively. Plot the
actual and estimated value of approximated function.

Sol:
Code written in ​Matlab
Considered Error metric as​ :MSE (Mean Square Error)
Outputs:
Plotting the function f(x) = 1+cos((pi/4)*x) in the range -5 to 5
Plotting the Actual And Estimated plot of the given function:

*********************************************************************************************************

Q2
: Implement the ART-1 learning algorithm to cluster six patterns [1, 0, 0, 0, 0], [1, 1, 0, 0, 0], [0,
0, 1, 0, 0], [0, 1, 1, 1, 0], [0, 0, 0, 1, 1] and [0, 0, 0, 0, 1] into three clusters. The following set of
parameter values are to be used: Vigilance parameter= .5, Learning parameter=2. Repeat the
experiment again for Vigilance parameter= .9, what changes did you observe?
Solution :

Code written in ​MATLAB


Results:

When ​vigilance parameter = 0.5


When vigilance parameter = 0.9

Observation:

Vi​gilance parameter helps in creating new memories or new


information higher the value the model will learn new details in
other case lower value will produce general memories.

You might also like