Glassdoor-Salary-Prediction: Overview

Collected, Cleaned and Analysed data from glassdoor.com
Performed Feature Engineering on various textual, categorical and numerical features in order to extract the significant areas to contribute to the model
Performed predictive modelling on the data using a variety of ML algorithms
Performed Hyperparameter Tuning to improve model performance
Evaluated the implemented machine learning models to get a MAE of 12.69
Implemented a Flask endpoint

Code and References

Python Version: 3.7
Packages: numpy, pandas, matplotlib, seaborn, selenium, pickle
Selenium Scraper: https://github.com/arapfaik/scraping-glassdoor-selenium
Code Reference: https://github.com/PlayingNumbers/ds_salary_proj.git
Productionization: https://towardsdatascience.com/productionize-a-machine-learning-model-with-flask-and-heroku-8201260503d2

Name		Name	Last commit message	Last commit date
Latest commit History 12 Commits
.ipynb_checkpoints		.ipynb_checkpoints
FlaskAPI		FlaskAPI
__pycache__		__pycache__
models		models
.DS_Store		.DS_Store
README.md		README.md
__init__.py		__init__.py
chromedriver		chromedriver
glassdoor_scraper.py		glassdoor_scraper.py
jobs_data.csv		jobs_data.csv
salary_prediction.ipynb		salary_prediction.ipynb