# 15. Logistic Regression¶

Todo

Zrobić aby wykorzystywało szablon _template.rst

## 15.1. Co to jest Logistic Regression?¶

In general when we make a machine learning based program, we are trying to come up with a function that can predict for future inputs based on the experience it has gained through the past inputs and their outputs (training set).

Logistic Regression is - coming up with a probability function that can give us ‘the chance, for an input to belong to any one of the various classes’ we have (classification).

Since the logistic function has two different asymptotes, it can be used to divide data into “yes/no” categories – the low side being “no” and the high side being “yes.”

Fig. 15.1. The standard logistic function ; note that for all .

## 15.2. Linear vs Logistics¶

Fig. 15.2. Linear vs Logistics

## 15.3. Podstawowe pojęcia¶

Binary Model
Model który ma dwa typy wartości (przykład: spam, nie spam)
logit
logistic function
logistic-sigmoid function
Funkcja sigmoidalna
Softmax function
takes logits and transforms them to probability distibutions

Todo

Bias term

Cost function

Entropy

Cross Entropy