New coder survey 2016

Predicting Developer Job Chances for Coding Bootcamp Participants

Download .zip Download .tar.gz View on GitHub

About

This project was my capstone project of the Udacity Machine Learning Engineer Nanodegree (https://www.udacity.com/course/machine-learning-engineer-nanodegree-by-google--nd009).

Project Summary

Predicted job chances for bootcamp participants after finishing a bootcamp based on a new coder survey published 2016 by FreeCodeCamp and CodeNewbie.org on kaggle.com.

Data

The "2016 New Coder Survey" can be found here: https://www.kaggle.com/freecodecamp/2016-new-coder-survey-.

Steps involved

  • preprocessed "2016 New Coder Survey" dataset
  • implemented training and testing data split
  • trained a random forest classifier to predict a bootcamp participants job chances
  • tweaked parameters of the classifier to reach a F1 score of 0.8832
  • visualized most important features that influence whether a bootcamp participant gets a job afterwards

Coding Image