# Linear Regression

Model:

Parameters:

## Square Loss

• Least Square Method
• Minimise Square Loss

## Normal Equation

Normal Equation: A method to solve for $\vec{\theta}​$ analytically

For m examples, n features:

## Comparison between Gradient Descent and Normal Equation

Gradient Descent Normal Equation
choose
many iterations
no iteration
works well even if n is large need to Compute
slow if n is very large