Model Representation

• Regression problem
  • Target variable to be predicted is continuous
• Classification problem
  • Target variable to be predicted is discrete

Cost Function

Accuracy of the hypothesis function. Takes an average difference of all results of the hypothesis with the input/output pairs.

Squared Error Function/Mean Squared Error

$$ h_\theta(x) = \theta_0 + \theta_1x

$$

$$ J(\theta0, \theta_1) = \frac{1}{2m} \sum^{m}{i=1}{(h_\theta(x_i) - y_i)^2} $$