Question

Question 1. Write a Matlab program to implement the following a. the logistic sigmoidal activation functions and b. tangential sigmoidal activation function Question 2. Calculate the output of the program for the following parameter a. Input values =−1 to 1 , with an interval of 0.001 , b. Lambda values =1 to 15 with an interval of 1 Question 3. Evaluate the output of the program by comparing the outputs and plotting Question 4. Write a program to implement the Mc-culloch and pitts neuron model for EX-or logic gate and verify the weight values and activation function.

Answer 1

The task is to write MATLAB programs for logistic and tangential sigmoid activation functions, run these with specified parameters, plot the results, and implement a McCulloch and Pitts neuron model for an EX-OR logic gate.

Step-by-step explanation:

The question pertains to the design and implementation of a MATLAB program that handles sigmoidal activation functions used in neural networks, as well as the Mc-Culloch and Pitts neuron model for creating an Exclusive-OR (EX-OR) logic gate. The question is split into several parts, requiring the programming of logistic and tangential sigmoid activation functions, testing these functions with sets of given parameters, and plotting the results for comparison. Moreover, the student is asked to implement the McCulloch and Pitts neuron model to simulate an EX-OR gate and verify the weight values and activation function.

An example of a MATLAB implementation for a logistic sigmoidal function would be:

function y = logistic_sigmoid(x,lambda)
y = 1./(1+exp(-lambda*x));
end

Similarly, an implementation for a tangential sigmoid function could be:

function y = tanh_sigmoid(x,lambda)
y = tanh(lambda*x);
end

Once the functions are defined, the next step is to evaluate their outputs over the given range of input values and lambda values, and plot the results for analysis. Implementing the McCulloch and Pitts neuron model for an EX-OR gate involves setting appropriate threshold values and weights to neurons in a way that mimics the EX-OR logic behavior.