HOUSE / TITANIC PREDICTION
APR 2026
Machine-learning prediction experiments using feature engineering and model evaluation across benchmark regression and classification datasets.
OVERVIEW
A machine-learning experimentation project covering predictive modeling on the House Prices (regression) and Titanic (classification) benchmark datasets, implementing end-to-end data cleaning, feature engineering, and cross-validated evaluation.
THE IDEA
Use classic benchmark datasets to implement and compare standard ML approaches end-to-end, evaluating how feature engineering directly impacts model performance.
WORKFLOW
KEY FEATURES
- Data preprocessing and missing-value imputation
- Categorical encoding and feature scaling
- Regression modeling on House Prices dataset
- Binary classification modeling on Titanic dataset
- Model training and cross-validation
- Evaluation metrics and error inspection
ENGINEERING APPROACH
Both datasets are processed through standard preprocessing pipelines tailored to regression and classification tasks respectively. Models are evaluated using cross-validation and standard evaluation metrics.