Recipe Rating Prediction with Imbalanced Data

In this project, I

performed modularized data cleaning processes on numerical and textual data based on food.com recipe data, ensuring data integrity for analysis.
conducted exploratory data analysis (EDA) and hypothesis testing to extract data insight and address the core research question.
built a comprehensive modeling pipeline, from data preprocessing, feature engineering, and model selection, to hyper-parameter tuning, increasing overall f1-score by 20% compared to the baseline model.