Interview Prep
Career
GenAI
Prompt Engg
ChatGPT
LLM
Langchain
RAG
AI Agents
Machine Learning
Deep Learning
GenAI Tools
LLMOps
Python
NLP
SQL
AIML Projects

Reading list

Machine Learning Basics for a Newbie
6 Steps of Machine learning Lifecycle Introduction to Predictive Modeling
Introduction to Exploratory Data Analysis & Data Insights Descriptive Statistics Inferential Statistics How to Understand Population Distributions?
Reading Data Files into Python Different Variable Datatypes
Probability for Data Science Basic Concepts of Probability Axioms of Probability Conditional Probability
Central Tendencies for Continuous Variables Spread of Data KDE plots for Continuous Variable Overview of Distribution for Continuous variables Normal Distribution Skewed Distribution Skeweness and Kurtosis Distribution for Continuous Variable
Central Tendencies for Categorical Variables Understanding Discrete Distributions Performing EDA on Categorical Variables
Dealing with Missing Values Understanding Outliers Identifying Outliers in Data Outlier Detection in Python Outliers Detection Using IQR, Z-score, LOF and DBSCAN
Sample and Population Central Limit Theorem Confidence Interval and Margin of Error
Bivariate Analysis Introduction
Covariance Pearson Correlation Spearman's Correlation & Kendall's Tau Correlation versus Causation Tabular and Graphical methods for Bivariate Analysis Performing Bivariate Analysis on Continuous-Continuous Variables
Tabular and Graphical methods for Continuous-Categorical Variables Introduction to Hypothesis Testing P-value Two sample Z-test T-test T-test vs Z-test Performing Bivariate Analysis on Continuous-Catagorical variables
Chi-Squares Test Bivariate Analysis on Categorical Categorical Variables
Multivariate Analysis A Comprehensive Guide to Data Exploration The Data Science behind IPL
Supervised Learning vs Unsupervised Learning Reinforcement Learning Generative and Descriminative Models Parametric and Non Parametric model
Machine Learning Pipeline Preparing Dataset Build a Benchmark Model: Regression Build a Benchmark Model: Classification
Evaluation Metrics for Machine Learning Everyone should know Confusion Matrix Accuracy Precision and Recall AUC-ROC Log Loss R2 and Adjusted R2
Dealing with Missing Values Replacing Missing Values Imputing Missing Values in Data Working with Categorical Variables Working with Outliers Preprocessing Data for Model Building
Understanding Cost Function Understanding Gradient Descent Math Behind Gradient Descent Assumptions of Linear Regression Implement Linear Regression from Scratch Train Linear Regression in Python Implementing Linear Regression in R Diagnosing Residual Plots in Linear Regression Models Generalized Linear Models Introduction to Logistic Regression Odds Ratio Implementing Logistic Regression from Scratch Introduction to Scikit-learn in Python Train Logistic Regression in python Multiclass using Logistic Regression How to use Multinomial and Ordinal Logistic Regression in R ? Challenges with Linear Regression Introduction to Regularisation Implementing Regularisation Ridge Regression Lasso Regression
Introduction to K Nearest Neighbours Determining the Right Value of K in KNN Implement KNN from Scratch Implement KNN in Python
Bias Variance Tradeoff Introduction to Overfitting and Underfitting Visualizing Overfitting and Underfitting Selecting the Right Model What is Validation? Hold-Out Validation Understanding K Fold Cross Validation
Introduction to Feature Selection Feature Selection Algorithms Missing Value Ratio Low Variance Filter High Correlation Filter Backward Feature Elimination Forward Feature Selection Implement Feature Selection in Python Implement Feature Selection in R
Introduction to Feature Engineering Feature Transformation Feature Scaling Feature Engineering Frequency Encoding Automated Feature Engineering: Feature Tools
Introduction to Naive Bayes Conditional Probability and Bayes Theorem Introduction to Bayesian Adjustment Rating: The Incredible Concept Behind Online Ratings! Working of Naive Bayes Math behind Naive Bayes Types of Naive Bayes Implementation of Naive Bayes
Understanding how to solve Multiclass and Multilabled Classification Problem Evaluation Metrics: Multi Class Classification
Introduction to Ensemble Techniques Basic Ensemble Techniques Implementing Basic Ensemble Techniques Finding Optimal Weights of Ensemble Learner using Neural Network Why Ensemble Models Work well?
Introduction to Stacking Implementing Stacking Variants of Stacking Implementing Variants of Stacking Introduction to Blending Bootstrap Sampling Introduction to Random Sampling Hyper-parameters of Random Forest Implementing Random Forest Out-of-Bag (OOB) Score in the Random Forest IPL Team Win Prediction Project Using Machine Learning Introduction to Boosting Gradient Boosting Algorithm Math behind GBM Implementing GBM in python Regularized Greedy Forests Extreme Gradient Boosting Implementing XGBM in python Tuning Hyperparameters of XGBoost in Python Implement XGBM in R/H2O Adaptive Boosting Implementing Adaptive Boosing LightGBM Implementing LightGBM in Python Catboost Implementing Catboost in Python
Different Hyperparameter Tuning methods Implementing Different Hyperparameter Tuning methods GridsearchCV RandomizedsearchCV Bayesian Optimization for Hyperparameter Tuning Hyperopt
Understanding SVM Algorithm SVM Kernels In-depth Intuition and Practical Implementation SVM Kernel Tricks Kernels and Hyperparameters in SVM Implementing SVM from Scratch in Python and R
Introduction to Principal Component Analysis Steps to Perform Principal Compound Analysis Computation of Covariance Matrix Finding Eigenvectors and Eigenvalues Implementing PCA in python Visualizing PCA A Brief Introduction to Linear Discriminant Analysis Introduction to Factor Analysis
Introduction to Clustering Applications of Clustering Evaluation Metrics for Clustering Understanding K-Means Implementation of K-Means in Python Implementation of K-Means in R Choosing Right Value for K Profiling Market Segments using K-Means Clustering Hierarchical Clustering Implementation of Hierarchial Clustering DBSCAN Defining Similarity between clusters Build Better and Accurate Clusters with Gaussian Mixture Models
Understand Basics of Recommendation Engine with Case Study
8 Ways to Improve Accuracy of Machine Learning Models
Introduction to Dask Working with CuML
Introduction to Machine Learning Interpretability Framework and Interpretable Models model Agnostic Methods for Interpretability Implementing Interpretable Model Understanding SHAP Out-of-Core ML Introduction to Interpretable Machine Learning Models Model Agnostic Methods for Interpretability Game Theory & Shapley Values
Introduction to Machine Learning Interpretability Framework and Interpretable Models model Agnostic Methods for Interpretability Implementing Interpretable Model Understanding SHAP Out-of-Core ML Introduction to Interpretable Machine Learning Models Model Agnostic Methods for Interpretability Game Theory & Shapley Values
Introduction to AutoML Implementation of MLBox Introduction to PyCaret TPOT Auto-Sklearn EvalML
Pickle and Joblib Introduction to Model Deployment
Deploying Machine Learning Model using Streamlit Deploying ML Models in Docker Deploy Using Streamlit Deploy on Heroku Deploy Using Netlify Introduction to Amazon Sagemaker Setting up Amazon SageMaker Using SageMaker Endpoint to Generate Inference Deploy on Microsoft Azure Cloud Introduction to Flask for Model Deploying ML model using Flask
Model Deployment in Android Model Deployment in Iphone

How to select Best Split in Decision Trees using Chi-Square

Himanshi Singh
Himanshi Singh Last Updated : 30 Mar, 2021
6 min read

Introduction

Welcome back! In the previous article, we learned about Gini impurity which we use to decide the purity of nodes. There is one more algorithm that we can use to decide the best split in decision trees and that algorithm is Chi-square.

Note: If you are more interested in learning concepts in an Audio-Visual format, We have this entire article explained in the video below. If not, you may continue reading.

What is Chi-Square?

Chi-square measures the statistical significance of the differences between the child nodes and their parent nodes. It is measured as the sum of squared standardized differences between observed and expected frequencies of target variable for each node and is calculated using this formula-
chi-square
Let’s see how we can calculate the expected values. If you recall this is how the split on “Performance in class” looks like-
chi-square Decision Trees
We’ve seen this before. There is a total of 20 students and out of those 10 play cricket and 10 do not. So, of course, the percent of students who do play cricket will be 50%. Now if we consider the “Above average” node here, there are 14 students in it, as the percentage of students who play cricket is 50% in the parent node as we discussed, the expected number of students who play cricket will of course be 7 and if you look at the actual value it is 8. So now we have both the values expected values and actual values.
The expected was 7 and the actual turns out to be 8, so we can infer that the expected value is calculated based on the distribution of the parent node of the same class. Similarly, the expected number of students who do not play cricket will be 7. I want you to intuitively think about this because remember the percentage of students who do not play cricket in the parent node is 50% as well and then here the actual value turns out to be 6.
Similarly for the below-average node expected to play and not play will be 3. Whereas the actual values are- 2 students play cricket and 4 do not.
Split performance in class decision trees
Now we can calculate the chi-square using the formula mentioned above for each child node. Can you guess what will be the chi-square value if the actual and expected values are the same? It’s actually pretty simple, it will be zero because both the actual and expected are the same and the difference will be zero.
Now if both values are the same we can generate an inference that the distribution of the child node is the same as there is the parent node and hence we are not improving the purity of the nodes. On the other hand, if the chi-square value is high it means that the distribution of child nodes is changing with respect to the parent node and we are going in a direction to
achieve more pure nodes hence we can say that:
Higher the chi-square value more will be the purity of the nodes after a split.

Properties of chi-square

Let’s look at some of the properties of chi-square before understanding how it actually works-
  • Chi-square just like Gini impurity works only with categorical variables so we cannot use it for continuous targets.
  • The higher the value of chi-square more the sub-nodes are different from the parent node and hence the homogeneity is more.
These are some of the properties of chi-square and you must consider these properties before choosing the right algorithm for deciding the split.

Steps to Calculate Chi-Square for a split-

Let’s now understand the steps to calculate chi-square for a split-
  • First, we need to calculate the expected values for each class.
  • Then we calculate the chi-square for individual nodes using this formula that we’ve seen before-
chi-square for decision trees
Here actual is what we actually have as the output and expected means what we’ve already calculated.
  • Finally, we calculate the chi-square for split using the sum of the chi-square of each child node for that split.
Don’t worry I’m going to show you an example and this will be extremely clear by the time you’re done with it so let’s again compare these two splits, one, of course, was on “the performance” and the other one was on “the class” they belong to and we’ll compare the chi-square statistic. This is the split on “performance in class” and the actual and expected values as we’ve already calculated earlier-
split of performance in class
Let’s now create a table to calculate the chi-square values for this split on “performance in class”:
So the actual values for the above-average node are shown in the rable. 8 students are above average and play cricket whereas 6 do not. The expected values will be 7 for both, as again we’ve discussed before. Similarly, we’ll calculate the values for the second node which will be the below-average node. So 2 students below-average play cricket and 4 do not and as a total of students below average only 6, expected values would be 3 both.
Next, we will calculate the deviation of actual values from the expected values for both classes:
calculate the deviation of actual values
So 8 – 7 will give us 1 and 6 – 7 will give minus 1. we will do the same thing for the below-average node as well. Finally, we have to calculate the chi-square values using the formula:
chi-square for best decision tree splits
We have (the actual – expected values) stored as deviations in the table. Remember these deviations are the actual – expected
values, we will square those values divided by the expected values and take the square root of those values. Let’s do that!
Above average Chi-Square(Play) = √ [(1)² / 7]  = √ 0.1428 ≈ 0.38
Below average Chi-Square(Play) = √ [(-1)² / 3]  = √ 0.3333 ≈ 0.58
So when you plug in the values the chi-square comes out to be 0.38 for the above-average node and 0.58 for the below-average node.
calculate the deviation of actual values
Finally the chi-square for the split in “performance in class” will be the sum of all these chi-square values:
performance in class
which as you can see here comes out to be 1.9 and this is the chi-square value for the split on “performance in class”.
Alright! similarly, we will calculate the chi-square value for the split on “the class” variable.
split on class decision trees
We will again calculate the table:
calculate table decision trees chi-square split
I suggest you calculate these values on your own and check your results with the values that you can see in this table. This will help you to understand how the chi-square process is working and finally, we’ll calculate the chi-square value for the split on class which will be the sum of all chi-square values:
new chi values
which comes out to be around 5.36. So what do you think we should do next? We will compare the two chi-square values and see which one is higher-
two chi-square values
And the population will split on “the class” variable as we can see here. Remember we saw the same result using the Gini index as well? The variable “Class” came out to be the split while using Gini Index as well.

End Notes

We saw the Chi-Square algorithm used for splitting the Decision Trees. This is also used for categorical targets. So we’ve covered two different algorithms so far and we saw that the results of both the algorithms have been quite similar.

If you are looking to kick start your Data Science Journey and want every topic under one roof, your search stops here. Check out Analytics Vidhya’s Certified AI & ML BlackBelt Plus Program

If you have any queries let me know in the comment section!

Decision TreesDecision Trees Chi-square
Himanshi Singh
Himanshi Singh

I’m a data lover who enjoys finding hidden patterns and turning them into useful insights. As the Manager - Content and Growth at Analytics Vidhya, I help data enthusiasts learn, share, and grow together. 

Thanks for stopping by my profile - hope you found something you liked :)

Free Courses

Generative AI 4.7

Generative AI - A Way of Life

Explore Generative AI for beginners: create text and images, use top AI tools, learn practical skills, and ethics.
Large Language Models 4.5

Getting Started with Large Language Models

Master Large Language Models (LLMs) with this course, offering clear guidance in NLP and model training made simple.
Prompt Engineering 4.6

Building LLM Applications using Prompt Engineering

This free course guides you on building LLM apps, mastering prompt engineering, and developing chatbots with enterprise data.
LlamaIndex 4.8

Improving Real World RAG Systems: Key Challenges & Practical Solutions

Explore practical solutions, advanced retrieval strategies, and agentic RAG systems to improve context, relevance, and accuracy in AI-driven applications.
RAG systems using LlamaIndex 4.7

Microsoft Excel: Formulas & Functions

Master MS Excel for data analysis with key formulas, functions, and LookUp tools in this comprehensive course.

Recommended Articles

Responses From Readers

Clear

Write for us

Write, captivate, and earn accolades and rewards for your work
  • Reach a Global Audience
  • Get Expert Feedback
  • Build Your Brand & Audience
  • Cash In on Your Knowledge
  • Join a Thriving Community
  • Level Up Your Data Science Game
imag
Congratulations, You Did It!
Well Done on Completing Your Learning Journey. Stay curious and keep exploring!
4 Simple Ways to Split a Decision Tree in Machine Learning (Updated 2024)
4 Simple Ways to Split a Decision Tree in Machine Learning (Updated 2024)
Decision Tree Algorithms – Simplified
Decision Tree Algorithms – Simplified

We use cookies essential for this site to function well. Please click to help us improve its usefulness with additional cookies. Learn about our use of cookies in our Privacy Policy & Cookies Policy.

Show details