Machine Learning No.3: Logistic Regression

1. Decision boundary

when h_θ(x) > 0, g(z) = 1; when h_θ(x) < 0, g(z) = 0.

so the hyppthesis is:

2. cost function

to fit parameters θ:

to make a prediction given new x:

Output       

3. Gradient Descent

Repeat {

　　

　　(simultaneously update all θ_j)

}