IPL Score Predictor

Project Overview

• Created a model that predicts the score (in terms of range) of IPL matches
• Optimized Multiple-Linear, Decision Tree, Random Forest, and AdaBoost regression models

How will this project help?

• This project is for the fantasy cricket fans out there, helping them to earn extra fantasy points for Dream11 IPL 2020

Resources Used

• Packages: pandas, numpy, sklearn, matplotlib, seaborn

Data Cleaning and Preprocessing

• Removing unwanted columns
• Keeping only consistent teams
• Removing the first 5 overs data in every match
• Converting the column 'date' from string into datetime object
• Handling categorical features

Model Building and Evaluation

Evaluation metric: Root Mean Squared Error (RMSE)
• Multiple Linear Regression - 15.843
• Decision Tree - 23.044
• Random Forest - 18.171
• Adaptive Boosting (AdaBoost) - 15.798

Future Scope

• Add columns in dataset of top batsmen and bowlers of all the teams.
• Add columns that consists of striker and non-striker's strike rates.
• Implement this problem statement using Artificial Neural Network (ANN).

Check out project in kaggle:

https://www.kaggle.com/swrnvh/first-innings-score-prediction-for-ipl