Software Automation
Logistic Regression

What is Logistic Regression?

Logistic regression is a supervised machine learning algorithm used to predict categorical outcomes, particularly binary classifications (e.g. yes/no, pass/fail, spam/not spam, cat/not cat).

While linear regression predicts numerical values, logistic regression outputs a probability between 0 and 1, which is then used to determine a class.

How Is Data Represented?

Features (input variables) are represented as binary values (0 or 1)

The target label is also binary (e.g., 1 for "cat", 0 for "not a cat")

Animal 4 Legs Whiskers Claws Is Cat

Animal 1 0 0 0 0

Animal 2 1 0 0 0

Animal 3 1 1 1 1

How Does It Work?

It uses a sigmoid function to calculate the probability that an input belongs to class 1.

𝑓(𝑥) = 1 / 1+𝑒⁻ˣ

The sigmoid function outputs values in the range (0, 1). A threshold (commonly 0.5) is used to classify:

• ≥ 0.5 → class 1 ("is a cat")

• < 0.5 → class 0 ("not a cat")

Example:

If the model outputs 0.84, we classify the sample as "cat".

Why Use Logistic Regression?

• It's simple, efficient, and interpretable

• Excellent for baseline binary classification tasks

• Can be extended to multi-class problems (multinomial logistic regression)

Extension Task:

Complete the Google Machine Learning Crash Course logistic regression module.