Linear Regression

Last Updated: 2021-11-19

Hypothesis

$$h_\theta(x) = \theta^T x$$

Parameter

$$\theta$$

Cost function:

$$J(\theta) = {1 \over 2m} \sum_{i=1}^m(h_\theta(x^{(i)}) - y^{(i)})^2$$

Goal

$$\underset{\theta}{\min} J(\theta)$$

Gradient Descent: simultaneously update all $\theta_j$

$$\theta_j:=\theta_j - \alpha {\partial \over \partial \theta_j} J(\theta)$$

Regularization

$$J(\theta) = {1 \over 2m} {\Big [} \sum_{i=1}^m(h_\theta(x^{(i)}) - y^{(i)})^2 + \lambda \sum_{i=1}^n \theta_j^2 {\Big ]}$$