the Titanic dataset, https://www.kesci.com/apps/home/dataset/58a940107159a710d916aefb. Actually, it's a classfication problem.
The projetc shows how to mine data, eg:process the original data then train the data with proper model.
- remove the outliers
- fill the null value
- analyse the original feature
- rebuild the future for training
- select proper model for training
- train data,save model, anlyse the feature importance