London Area, United Kingdom
• Cleaned and preprocessed a large-scale wastewater microbiome dataset; performed feature engineering guided by biological domain knowledge.
• Trained and evaluated 15+ machine learning models to (1) classify wastewater microbiomes, (2) predict treatment efficiency indicators, and (3) forecast microbiome composition under different operating conditions.
• Conducted comparative performance analysis across models to identify approaches that best support wastewater treatment plant efficiency.
• Applied SHAP for interpretability, analyzing feature–target relationships and assessing how microbiome features may contribute to treatment efficiency.
• Built a reusable ML code package (21 models) including standardized preprocessing, hyperparameter tuning, and visualization modules to support repeatable experimentation.
• Co-authoring a paper on Machine Learning in Wastewater Treatment with Jiangwen Dong, David Jiang, Tom Vinestocka, and Miao Guo (manuscript in preparation).
