Titanic Survival Predictions (Python)

A binary classification model, inspired by the "Titanic" Kaggle Challenge.

Predicts whether or not a given passenger will survive, based on personal characteristics such as age, gender, and how much money their ticket cost.

Data Dictionary

Setup

Prerequisites:

• Anaconda and Python 3.7
• Graphviz (brew install graphviz)
• Orca (conda install -c plotly plotly-orca)

Setup virtual environment:

conda create -n titanic-env python=3.7
conda activate titanic-env

Install package dependencies:

pip install -r requirements.txt

Usage

Import the data, generate profile reports, and train and score the classifier:

python -m app.importer
python -m app.profiler
python -m app.classifier

Results

Feature Importances:

Decision Tree Logic: