kaggle-titanic and How-to-score-0.8134-in-Titanic-Kaggle-Challenge
These are complements: the tutorial provides foundational techniques for data munging and supervised learning that form the basis for implementing the advanced feature engineering and model optimization strategies demonstrated in the competition solution.
About kaggle-titanic
agconti/kaggle-titanic
A tutorial for Kaggle's Titanic: Machine Learning from Disaster competition. Demonstrates basic data munging, analysis, and visualization techniques. Shows examples of supervised machine learning techniques.
Implements a complete competitive analysis workflow in a single IPython Notebook, covering feature engineering, exploratory data analysis with Matplotlib visualizations, and model comparison across multiple algorithms (logistic regression, SVM with multiple kernels, random forests). Leverages the PyData stack—NumPy, Pandas, scikit-learn, and StatsModels—to demonstrate k-fold cross-validation for local evaluation and direct submission to Kaggle's competition API. Includes benchmark reference scripts to help newcomers understand foundational approaches to the prediction task.
About How-to-score-0.8134-in-Titanic-Kaggle-Challenge
ahmedbesbes/How-to-score-0.8134-in-Titanic-Kaggle-Challenge
Solution of the Titanic Kaggle competition
Implements a complete machine learning pipeline including exploratory data analysis, feature engineering, and ensemble model tuning on structured passenger data. The solution employs data cleaning, feature transformation, and hyperparameter optimization techniques to achieve 0.8134 accuracy. Delivered as an interactive Jupyter notebook documenting each stage of the predictive modeling workflow.
Scores updated daily from GitHub, PyPI, and npm data. How scores work