This is a beginner competition challenge from Kaggle, where I put together my statistical analysis techniques, machine learning knowledge and Python skills to achieve more accurate predictions on who survived the Titanic disaster.
I believe everyone knows about Titanic from the famous movie to the true tragedy story. However, I've never thought that it can be a topic for machine learning. When I found this project on Kaggle, I was really excited and hit the join button without hesitation.
Out of 2224 passengers and crew on board, only 722 people survived. Despite luck in their surviving, there are more to be discovered. With passenger information, such as name, sex, ticket number, cabin class, family numbers and etc., factors behind the survival are demystified by various data-related techniques.
- Create a model to accurately predict whether a passenger will survive or not
- Answer the question: what sorts of people were more likely to survive?
- Python for
- data pre-processing
- exploratory data analysis
- statistical analysis
- machine learning model building
- Statistical analysis for feature engineering and model insights
- Mutual Information Analysis
- Principal Component Analysis
- TensorFlow Random Forest Model with automated hyper parameter tuning
- tfdf.keras.RandomForestModel
The data is published on Kaggle.
- Kaggle: Beginner Titanic Challenge