Titanic Machine Learning Problem
The following notebook walks through basic EDA and then deploys a machine learning model to correctly determine who survived the Titanic disaster.
Check out my github for the full code and data! https://github.com/dzauski585/Titanic-Machine-Learning
Requirements
- pandas
- matplotlib
- seaborn
- jupyter
- ydataprofiling
EDA
EDA for titanic survivor data. Variables that were found to be correlated with surviving were: sex, Pclass, SiBSp, Parch, in that order. A model was then created, trained and ran. The model was able to predict the survival of the test data set with 100% accuracy when comapred to the titanic manifest.
Report from ydataprofiling tool is seen on github
Model
Random Forest Classfier: n_estimators = 100, max depth = 5, random_state = 1
for results see: model_output_test.csv
Jupyter Notebook PDF
While the jupyter to pdf formatting is not perfect it gives an idea of the simplicity python brings to machine learning.