📊 A comprehensive comparison of TabNet and XGBoost across binary classification, multiclass classification, and regression tasks, showcasing performance metrics and fine-tuning results.

Employ the Spain's European Health Survey to predict risk of depression/anxiety

PLAYER RATING ANALYSIS

Classifying whether an asteroid is hazardous or not.

Data science project predicting SuperStore sales with Linear Regression and XGBoost. Uses date-based, engineered (discount, competitor price), and encoded categorical features. Includes preprocessing, MSE evaluation, and visualizations. Part of #60DaysOfLearning2025.

Timeseries analysis - Study case of a small business using ARIMA,SARIMA and XGBoost models to predict stock based on sales.

Feature attribution for realized model fit (rather than predictions). EDEF provides exact additive decomposition of predictive performance for linear, PyTorch, and tree-based models, including standard errors, grouped inference, and exact TreeIG decompositions for tree-based models.

A Machine Learning–based credit card fraud detection system using a hybrid ensemble of Random Forest and XGBoost. and LightGBM and CatBoost The project handles highly imbalanced data using SMOTE, performs and real-time fraud transactions prediction

Predicting residential energy demand for July using Random Forest and XGBoost in R. Live Shiny app deployed on shinyapps.io.

A loan default payment model

Short analysis of the UCI heart disease analysis, as well as walking through building a predictive model gradient boosted regression model.

📊 Predicting telecom customer churn to enable targeted retention campaigns — XGBoost & feature engineering.