Want to create interactive content? It’s easy in Genially!
Intro to Decision Trees
Sive Lowell
Created on March 13, 2025
Start designing with a free template
Discover more than 1500 professional designs like these:
View
Smart Presentation
View
Practical Presentation
View
Essential Presentation
View
Akihabara Presentation
View
Pastel Color Presentation
View
Visual Presentation
View
Relaxing Presentation
Transcript
An introduction to
decision trees
for classification
START
overview
Decision trees are an important tool in Machine Learning and Artificial Intelligence. They utilize features of a dataset to classify new datapoints. By the end of this lesson, you will understand the logic behind decision trees and practice using them to classify new instances of a datapoint.
Image from mungfali.com
Index
Decision Trees Overview
Practical Uses
Practice!
Decision trees are a way to model decision-making in a tree-like structure, where regular nodes are questions and leaf nodes are decisions. By traversing the tree starting at the root, we can make a decision, whether it's a yes/no problem or a more complicated issue of classification.
What are decision trees?
Key concepts
Entropy is a measure of the uncertainy of a variable. For example, flipping a coin has high entropy because it is uncertain if it will be Heads or Tails.
Entropy
Information gain tells us how much new information is added by a variable. In other words, it if a variable gives us a better understanding of our data, it has high information gain.
Information Gain
The best split feature is the feature that reduces entropy the most. If a certain feature helps us guess how to classify an instance of our dataset, we want to use it early in our tree.
Best Split Feature
More on information theory
In ML, decision trees can be especially useful for classification. When we are given an example of an object, we can traverse our decision tree and ask questions at each node that help us get closer to identifying what our new example is.