COP701 - Toy Neural Network - IT Assignment Help

Assignment Task

In this assignment, you are expected to submit the following:
1. Implement a multi-layer perceptron (MLP) based neural network using C.
2. The implementation should be a parameter-based library. The client or user can define the activation function, cost function and backpropagation technique using some arguments and then build the .so file. The activation functions can either be Sigmoid or Tanh, or ReLu. The backpropagation algorithm can be either Batch gradient descent or stochastic gradient descent or mini-batch gradient descent. The cost function can be either cross-entropy or mean squared error
3. Implement an MLP classifier using Scikit learn library in python.
4. Design multiple multilayer perceptrons with different activation functions. Plot a loss function graph, wherein the x-axis would be your number of iterations, and the y-axis would be the loss function value with every iteration.
5. The loss function graph should have multiple curves corresponding to the Scikit-learn version of MLP, your implementation of the MLP classifier with different activation functions such as Sigmoid or Tanh, or ReLu. You can have two versions of the loss function graph, one with cross-entropy as a cost function and one with mean squared error as a cost function.
6. Record the accuracy of the classifier and also state the observation in the report.
7. Simulate gradient vanishing problem using a suitable cost function, activation function and backpropagation algorithm. Can we get rid of gradient vanishing problems? If yes, how and If no, why?

This COP701 - IT Assignment has been solved by our IT experts at My Uni Paper. Our Assignment Writing Experts are efficient to provide a fresh solution to this question. We are serving more than 10000+ Students in Australia, UK & US by helping them to score HD in their academics. Our Experts are well trained to follow all marking rubrics & referencing style.