š Hi, Iām Manas. I'm a graduate student at Pace University. I am passionate about data. My goal is to display my work and document the process of EDA and ML models that I build. The github pages will also serve as a way for me to reflect/revisit my past work. For job opportunities feel free to reach out at - [email protected]
- -> Chocolate Hackathon folder is the work done for the DataIku hackathon. My team ended up 3rd out of 35 teams. We were given a weekend to come up with a predictive model that can predict chocolate ratings from chocolate focus group dataset. We used hyperparameter tuning and random forest regressor to predict the ratings with an accuracy of 85.6%. In retrospective, my team did not use sklearn's pipeline module to make models. Using pipelines, would have significantly reduced the time spent on Individually making a model.
- -> 4 ML models with EDA