Hanan User

I’m an aspiring Data Scientist with hands-on experience in machine learning, predictive analytics, and NLP using Python. I enjoy taking ideas from raw data to reliable models—building end-to-end ML pipelines that cover data preprocessing, feature engineering, model training, evaluation, and optimization.

In my projects, I’ve focused on predictive modeling that supports real-world decisions. For customer churn prediction, I evaluated whether price sensitivity was a key business driver for an energy utility, using exploratory data analysis, class imbalance handling (SMOTE), threshold optimization, and model development with Random Forest and XGBoost—then translated the results into retention recommendations and clearer signals of churn risk.

I also built a multi-task NLP system using DistilBERT to predict driver intent, emotional state, and optimal driving mode from natural language commands. I designed and trained a custom multi-head architecture, implemented an end-to-end NLP pipeline (dataset preparation, preprocessing, tokenization, label encoding, and transformer fine-tuning), and improved training stability using class-weighted loss functions with AdamW optimization and learning-rate scheduling. I’m motivated by work that pairs strong modeling rigor with practical impact, and I’m ready to contribute where learning, collaboration, and measurable outcomes matter.