​

The course is an invitation to use decision trees for binary classification problems,

The course is divided into classes as follows:
1. What is machine learning and what are decision trees?

2. A look at Python.

3. Formal description of algorithms that learn decision trees
4. Implementation of decision trees for churn rate prediction. 5. Advanced aspects of decision trees (three hours).

​

​

Syllabus

​

1. Introduction

   • What is Machine Learning and Decision Trees ?

   • Mathematical notation

2. Python installation and a working environment for MachineLearning

   • Mini-microwave installation

   • Installation for Windows.

   • Installation for MacOS

   • Installation for Linux

   • Creation of a virtual environment for data analysis

   • Disable the base environment

   • Creation of a specific environment for data analysis using conda-forge

3. Decision trees and their training algorithms

   • Decision trees

   • Learning algorithm and the concept of entropy

4. Churn rate binary classifier for telecommunications companies

   • Churn rate

   • Database Description

5. Selected topics in decision trees

   • Motivation of the concept of entropy

   • Gini impurity index

   • Random forests vs over-fit

   • Pruning

​