Click on the title to read more about the project.

PROJECT 01

Tennis Win Prediction

Developed a model to predict a tennis player’s win probability based on the competing players’ match performance statistics and player characteristics. Used ensemble technique to improve performance. Deployed the model through an app using streamlit.io

Photo by cottonbro on Pexels.com

PROJECT 02

Job Title & Salary Prediction

Scraped local job search platform for data science job ads. Used natural language processing techniques to develop a model that predicts job title and salary bands based on the job ad details. Compared performance of two types of vectorizers and determined the best vectorizer-model combination. Compared performance of supervised and unsupervised models.

Photo by Mikael Blomkvist on Pexels.com

PROJECT 03

Property Sale Price Model

Used the Ames Housing Data set to develop a regression model to predict property sale price based on non-renovatable property characteristics. Determined how the property price is affected by its renovatable features. Used classification techniques to identify factors that result to an abnormal sale.

Photo by David McBee on Pexels.com