debajyoti94 / myheartwillnotgoon

Finding out who survived on Titanic ship using Logistic Regression.

Who will survive on Titanic?

Building a predictive model to see who will survive on titanic ship. Dataset is taken from here.

The repository has multiple directories, with each serving a different purpose:

• input/: contains the dataset, split into train and test files.
• model/: consists of baseline logistic regression models. Stratified Kfold validation was applied while training the model, with number of folds=5, hence you see 5 models.
• notebooks/: Consists of one jupyter notebook. It was used for EDA purpose and also experiment with some functions used for feature engineering.
• src/: this directory consists of the source code for the project.
  • config.py: consists of variables which are used all across the code.
  • create_folds.py: used for implementing stratified kfold cross validation.
  • feature_engg.py: used for cleaning the data (test & train) and apply data imputation for missing values in the features.
  • test_functionalities: using pytest module, i define some data sanity checks on the training data.
  • train.py: this file contains the code for implementing the model. The train and the inference stage.

To train the model use the following command:

python train.py --train skfold

For inference stage, use:

python train.py --test inference