This is the legendary Titanic ML competition, designed as the best first challenge for those looking to dive into machine learning competitions and familiarize themselves with how the Kaggle platform works. It is a "Getting Started" competition created by Kaggle data scientists for individuals with little to no machine learning background.
The sinking of the RMS Titanic on April 15, 1912, remains one of the most infamous shipwrecks in history. While there was an element of luck involved in surviving, certain groups of people were more likely to survive than others. In this challenge, participants are asked to build a predictive model that answers the question: "What sorts of people were more likely to survive?"
Using passenger data—including name, age, gender, socio-economic class, and other details—competitors must predict survival outcomes.
Participants gain access to two datasets: - train.csv: Contains details for 891 passengers, including whether they survived (the "ground truth"). - test.csv: Contains similar information for 418 passengers but does not disclose the survival outcome. Participants must use patterns found in the training data to predict these outcomes.
Submissions must be a CSV file with exactly 418 entries plus a header row. The file must contain exactly two columns:
- PassengerId (sorted in any order)
- Survived (binary predictions: 1 for survived, 0 for deceased)
Note: These competitions have no cash prize and operate on a rolling timeline.