Pinned Repositories
food-desert-prediction
Capstone Project: Pulled tweets from Twitter API to incorporate sentiment and population health into a predictive model of food deserts with an AUC of 0.88 and accuracy of 0.89. Built an interactive Flask web app hosted on AWS using Carto to map health trends across the U.S.
fraud-detection
Detected fraud with AUC of 0.94 and f1-score of 0.91 using Random Forest. Live event details predictions were stored in MongoDB and integrated into a Flask web app hosted on AWS.
mlflow
Open source platform for the machine learning lifecycle
movie-recommender
Utilized recommender systems to predict user movie ratings. Ran Spark ALS Matrix Factorization to predict movie ratings with an average rating of 4.32/5 for the top 5% of movies.
ridershare-churn-prediction
Predicted rideshare customer churn with an accuracy of 0.79 using gradient boosting. Profit curve analysis according to business need yielded $68,000 through promotion suggestions.
karinapatel's Repositories
karinapatel/food-desert-prediction
Capstone Project: Pulled tweets from Twitter API to incorporate sentiment and population health into a predictive model of food deserts with an AUC of 0.88 and accuracy of 0.89. Built an interactive Flask web app hosted on AWS using Carto to map health trends across the U.S.
karinapatel/fraud-detection
Detected fraud with AUC of 0.94 and f1-score of 0.91 using Random Forest. Live event details predictions were stored in MongoDB and integrated into a Flask web app hosted on AWS.
karinapatel/mlflow
Open source platform for the machine learning lifecycle
karinapatel/movie-recommender
Utilized recommender systems to predict user movie ratings. Ran Spark ALS Matrix Factorization to predict movie ratings with an average rating of 4.32/5 for the top 5% of movies.
karinapatel/ridershare-churn-prediction
Predicted rideshare customer churn with an accuracy of 0.79 using gradient boosting. Profit curve analysis according to business need yielded $68,000 through promotion suggestions.