This repository explores Univariate and Multivariate Linear Regression, as well as the Normal Equation Method of solving an equation.
You can get in touch with me on LinkedIn!
If you liked my repository, kindly support my work by giving it a ⭐!
The cost function of Linear regression is given as:
The parameter(s) theta, are updated with the equation below.
Notice that alpha is the learning rate/step size.
The equations can be vectorized and implemented as matrices.
This is an effective tool for debugging an algorithm.
The cost MUST go down over the number of increasing iterations.
This method is an alternative to regression, and solves for the parameters directly.