- Define two distributions as two classes.
- Sample 1000 points from two distributions and define their class labels.
- Create a linear classification model. Initialize random weights and plot samples and classification boundary
- Optimize weights using stochastic gradient descent (LMS) learning algorithm for least mean squared error.
- Compare initial classification boundary with final (optimized) classification boundary
- Plot learning curve (MSE vs epochs)
- Plot sigmoid function and it's derivative with-respect to stimulus 'x'