Documenting my hackathon projects, Kaggle competitions, and experiments, including wins, baselines, and approaches I tried along the way. These projects reflect my participation in data challenges, from classic Kaggle problems (Titanic, Disaster Tweets) to domain-specific datasets (earthquake damage, steel plate defects, loan approvals, etc.).
- 2024 Loan Approval – ML models for predicting loan eligibility.
- 2024 Regression with a Flood Prediction Dataset – Time-series/regression for flood impact prediction.
- 2024 Regression with an Abalone Dataset – Predicting abalone age from physical measurements.
- 2024 Santa: Perplexity Permutation Puzzle – Kaggle seasonal puzzle competition.
- Disaster Tweets – Classifying tweets as real or not about disasters.
- Predicting Earthquake Damage – Severity prediction for buildings after an earthquake.
- Steel Plate Defect – Detecting types of steel defects.
- Titanic Competition – The classic ML starter challenge.
- Practice end-to-end ML pipelines: EDA → feature engineering → modeling → evaluation.
- Apply different algorithms (linear models, trees, ensembles, neural nets).
- Document lessons learned from each competition.