• Built a sentiment analysis model in Python employing Neural Networks in Keras to categorize bookreviews for emotional tone, achieving a precision of 85%.• Applied one-hot encoding to preprocess the categorical Airbnb host data, used winsorization toreduce the outlier effect, and integrated logistic regression to predict the host ratings, achieving aprecision of 87%.• Lead a team of 5 to train machine learning models on 100K+ training data to predict fraudulenttransactions for Accenture. Build models using binary classification and logistic regression. Utilize upsampling to solve the class imbalance problem.

Led interactive worksheets, conducted informative office hours, and delivered engaging lab previews to over 300 students, helping students better understand course concepts and achieve academic excellence in “CS 33: Introduction to Computer Organization”.