1. Predictions of mortality in ICU patients
· Performed EDA and feature selection to identify factors affecting patient survival.
· Applied logistic regression and tree-based models to predict mortality risk.
2. Rossman drug stores forecasting sales
· Collected and cleaned sales data across 1,115 stores
· Built predictive models (XGBoost, LightGBM etc.) to forecast daily sales, achieving high model accuracy.
3. Predicting student performance in virtual internships
· Engineered team-level features and analysed chat data with NLP(TF-IDF)
· Developed predictive models for student performance evaluation
4. Social media analysis for World-level Events: World Cup 2022
· Gathered and analysed social media data from Twitter, Facebook, Instagram
· Performed sentiment analysis and topic modelling (LDA) to extract key engagement trends.
5. Tesla revenue growth forecast
· Applied ARIMA, SARIMA, Random Forest, and XGBoost etc. on historical data.
· Conducted hyperparameter tuning (GridSearchCV) and multivariate regression modelling to forecast quarterly revenue.
Tennis | Dance (Hip-hop, Jazz, K-pop) | Playing Guzheng(Chinese traditional instrument) | Badminton