In this project, I have implemented Maximum Likelihood Estimation, Maximum Priori Estimation for regression. I also show the improvement with regularization.
Functions used:
- X = hypothesis func(data,order) This function is going to generate a (N X order+1) data matrix, where N is the length of data.
- [x,y,t] = generateDate(mu,sigma,npts) This function is going to generate the data points x, desired values without gaussian noise y and the target t for a given mean mu, standard deviation sigma and the total number of data points npts.
The code is divied into 4 sections. It is generalized. So, you can change the order vector, the number of points, lambda, beta and alpha values according to need. The code is generic. Order, lamda and alpha vectors can be of any length. Enter the desired values accodingly.