Multilayer Perceptron (MLP)

Implementation of multilayer perceptron/neural network from scratch to understand and grasp basic concepts of Deep Learning.

Requirements:

1. numpy
2. matplotlib
3. scikit-learn

Output plots:

1. Generated dataset for classification

2. Decision boundary from Logistic Regression model

3. Decision boundary from MLP with hidden layer size = 3

4. Variation in decision boundary with hidden layer size

This exercise shows the capability of neural network to form non-linear decision boundaries where logistic regression fails through an example. Also, shows the variation in decision boundary as we increase the number of nodes in hidden layer.