Skip to content

deedGhost/DonorsChoose

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 

Repository files navigation

DonorsChoose Project Approval Prediction

Overview

This project involves building machine learning models to predict whether a project proposal on the DonorsChoose platform will be approved for funding. The dataset used for this analysis was sourced from the Kaggle. The project explores various models, including Logistic Regression, Decision Tree, Random Forest, and XGBoost, to identify the key factors that influence project approval.

Dataset

The dataset is sourced from the Kaggle. It contains features related to teachers, schools, projects, and project descriptions, which are used to predict whether a project will be approved.

Modeling

The project explores several machine learning models:

  • Logistic Regression: Used both with and without hyperparameter tuning.
  • Decision Tree: A tree-based model that helps understand feature importance.
  • Random Forest: An ensemble method that averages multiple decision trees to improve accuracy.
  • XGBoost: A powerful gradient boosting algorithm used with hyperparameter tuning.

Observations

  • Logistic Regression: Provided a strong baseline with decent accuracy, but struggled with complex patterns. TEST ACCURACY-85%
  • Decision Tree: Overfit on the training data but gave insights into feature importance. TEST ACCURACY-75%
  • Random Forest: Improved generalization over Decision Tree but required more computation. TEST ACCURACY-85%
  • XGBoost: Outperformed other models, particularly after hyperparameter tuning, showing strong predictive power. TEST ACCURACY-85%

Results

The models were evaluated using metrics like accuracy and AUC (Area Under the Curve). The best model was selected based on cross-validation performance and its ability to generalize to unseen data.

  • Best Model: XGBoost with hyperparameter tuning achieved the highest AUC score.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published