/Titanic-Challenge

My take on the Kaggle Titanic Challenge, Accuracy: 0.80681

Primary LanguageJupyter NotebookMIT LicenseMIT

Titanic-Challenge

My take on the challenge Titanic: Machine Learning from Disaster

Preprocessing and Feature Engineering

  • Dropped 'Ticket' and 'Cabin' features
  • Made a new feature 'Title' by using the 'Name' feature.
  • Made a new ordinal feature 'Family_size' using 'Parch' and 'SibSp'.

Model and Accuracy

  • Endemble of 11 Random Forest Classifier models for final submission.
  • Accuracy : 80.861 (Public Leaderboard)