HR Employee Attrition Analysis and prediction

Real-world dataset with 35+ HR features

Extensive EDA, feature engineering & preprocessing pipelines

Classical ML models: Logistic Regression, Random Forest, AdaBoost, XGBoost

Hyperparameter tuning via RandomizedSearchCV

? Custom Deep Neural Network built with PyTorch (500 epochs, ReLU, dropout, Adam optimizer, BCELoss)

DNN outperformed traditional models in generalization

Deployed with Streamlit on Azure Web Services

️ Tech Stack

Python, Pandas, Scikit-learn, PyTorch, XGBoost, Streamlit, Azure App Services