Kaggle Titanic Machine Learning

by wyattwang7

GitHub Readme.md

Kaggle Titanic: Machine Learning Modeling and Deployment

This repository builds a machine learning model on Titanic dataset, creates a web app with Flask to predict the probability of survival per passenger, and deploys the app to Heroku.

Quick Look

  • This video will help you get a feel for the app.
    .

  • Also you could try the app yourself by clicking the button below. You might experience pretty low latency due to the policy of Heroku.
    Deploy

Keywords

EDA, custom estimator, tree classifiers, boosting, model persistance, html, flask, Heroku

Prerequisites

  • Datasets
    Here is the overview of the project.
  • Dependencies
    xgboost==0.90
    Flask==1.1.1
    dill==0.3.0
    numpy==1.16.4
    requests==2.22.0
    pandas==0.25.1
    scikit_learn==0.21.3
    gunicorn==19.9.0

Accomplishment

  • EDA, feature engineering and modeling (Titanic.ipynb)
  • Model deployment (#clone the repository and run locally)
    1. export FLASK_APP=app.py
    2. python -m flask run
    3. Check the link in the terminal