https://www.linkedin.com/in/assylgul-janayeva-19547720b?utm_source=share&utm_campaign=share_via&utm_content=profile&utm_medium=ios_app Projects Geo-aware ML model for Methane concentration prediction (Dissertation project, Cardiff University) • Collected and merged multimodal environmental datasets (CH₄ measurements, meteorological data, geospatial metadata). • Implemented and compared multiple ML models: Random Forest, XGBoost, Gradient Boosting Regression (GBR), SVR. • Applied feature selection (variance threshold, mutual information) and SHAP for interpretability of key influencing factors. • Evaluated performance with RMSE, MAE, and R², demonstrating improved generalization to unseen stations. Factual vs. Opinion Classification (NLP, Cardiff University) • Built an NLP model to classify news articles as factual or opinion. • Used TF-IDF and ML models (SVM, Logistic Regression), achieving 98% accuracy. • Applied lemmatization, stopword removal, and POS tagging to improve text preprocessing. • Collaborated in a student team using Agile and Scrum practices for sprint planning, backlog management, and iterative model development. Hotel Booking Analytics Dashboard (Data Visualization, Cardiff University) • Processed and analyzed 100k+ booking records using Python and SQL. • Built an interactive Tableau dashboard with KPIs on cancellations, revenue loss, and customer segmentation. • Delivered insights for revenue forecasting and cost optimization, supporting managerial decision- making.