10.18453/rosdok_id00005013 Fedorov, Aleksandr Aleksandr Fedorov http://d-nb.info/gnd/1382845707 https://orcid.org/0000-0001-6434-6623 Universität Rostock 2024 004 Data processing Computer sciences 540 Chemistry & allied sciences 660 Chemical engineering 2024 en https://purl.uni-rostock.de/rosdok/id00005013 urn:nbn:de:gbv:28-rosdok_id00005013-7 The present work focuses on applying modern data science and machine learning (ML) methods to investigate CO2 hydrogenation to higher hydrocarbons, also known CO2-Fischer-Tropsch synthesis (CO2-FTS). These methods were used for literature analysis on CO2-FT catalysts and for developing kinetic models with neural networks. New data normalization approaches and improved ML models, incorporating chemical and chemical engineering knowledge, were developed to handle limited and small data.