Projekt_mlynatom package provides data processing of Titanic - Machine Learning from Disaster data and prediction of survival by logistic regression and neural networks.
The package is not available from official repositories and can be installed with the following command.
(@v1.8) pkg> add https://github.com/B0B36JUL-FinalProjects-2022/Projekt_mlynatom| Variable | Definition | Key |
|---|---|---|
| survival | Survival | 0 = No, 1 = Yes |
| pclass | Ticket class | 1 = 1st, 2 = 2nd, 3 = 3rd |
| sex | Sex | |
| Age | Age in years | |
| sibsp | # of siblings / spouses aboard the Titanic | |
| parch | # of parents / children aboard the Titanic | |
| ticket | Ticket number | |
| fare | Passenger fare | |
| cabin | Cabin number | |
| embarked | Port of Embarkation | C = Cherbourg, Q = Queenstown, S = Southampton |
pclass: A proxy for socio-economic status (SES) 1st = Upper 2nd = Middle 3rd = Lower
age: Age is fractional if less than 1. If the age is estimated, is it in the form of xx.5
sibsp: The dataset defines family relations in this way... Sibling = brother, sister, stepbrother, stepsister Spouse = husband, wife (mistresses and fiancés were ignored)
parch: The dataset defines family relations in this way... Parent = mother, father Child = daughter, son, stepdaughter, stepson Some children travelled only with a nanny, therefore parch=0 for them.
This package focuses on two main parts: Data processing and prediction of survival. All functions are implemented in files in src folder. Example usage is shown in classify.jl.
In this part (mainly in src/data_preparation.jl) are data processing functions provided. Main focus is on processing categorical values, filling missing values and creating new data from other.
- categorical values are converted to dummy encoding
- missing age, fare and embark are filled with data, see classify.jl.
- from name column titles are separated and then used as new data
In this part are implemented 2 options. First is ridge logistic regression with normal/adam step and second are neural networks.
- Neural networks are trained and defined using Flux library with custom training loop.
- Ridge logistic regression is regularization method used to avoid overfitting.
- Adam optimization algorithm is used for better (and faster than in normal gradient descent).