报告时间：2025年11月9日（星期日）16:20-16:50
报告地点：合肥明珠瑞士大酒店瑞士厅
报 告 人：Jonas Kristanto
工作单位：印度尼西亚加查马达大学
举办单位：机械工程学院
报告简介：
Biomass thermochemcial conversion is an important pathway to produce green carbon-based energy and sustainable material. The presentstion focuses on the development of biomass pyrolysis kinetics with the aid of deep machine learning approach. The talk will detailedly introduce implementation of machine learning techniques using Linear Regression, Random Forest, and Gradient Boosting regressors with enhanced preprocessing that incorporates feedstock taxonomy information such as genus and family as boolean parameters. The addition of taxonomy rank information increases model performance by up to 60% for Linear Regression and about 6% for both Random Forest and Gradient Boosting. Feature importance analysis identified temperature, volatile content, and hydrogen concentration as dominant factors. In conclusion, integrating taxonomy-based preprocessing with machine learning substantially improves the reliability and interpretability of biomass pyrolysis yield prediction.
报告人简介：
Jonas Kristanto is a faculty member in the Department of Chemical Engineering, Faculty of Engineering, Universitas Gadjah Mada. He serves as an assistant professor specializing in biomass thermochemical conversion, with a particular focus on pyrolysis kinetics and modeling. Jonas earned both his bachelor’s and doctoral degrees in chemical engineering from Universitas Gadjah Mada, completing his Ph.D. in 2023. His current work involves various industrial collaborations in digitalization of chemical processes, life cycle assessment, and other environmental engineering topics. These collaborations allow him to integrate practical industrial insights into his research and teaching activities.
