gigilui15 / cis2350---assignment

MLP to learn the 5-input XOR problem (Without using ML Libraries)

CIS2350---Assignment

MLP to learn the 5-input XOR problem (Without using ML Libraries)

'XORProblem.py' is a script of the 'Assignment' MLP which had the following requirements:

Implement a multi-layer perceptron (MLP) to learn the 5-input XOR problem. The model is built as demonstrated in class. You must implement the project in Python without using AI and ML libraries such as Keras. Use a (random) 80/20 training/test split. Use sigmoid for the hidden and output layer neurons. Use a single output for the class label (0 to 1). Your MLP must therefore have 5 input and 1 output neurons. Start with 4 hidden layer neurons and then experiment with different values. Express results as simple classification error. Discuss, and interpret, the results. Extras - what other experiments, or visualizations, can you do? You do not need a GPU to train the network.