%0 Journal Article %1 panagiotougenerative %A Panagiotou, Emmanouil %A Qian, Han %A Marx, Steffen %A Ntoutsi, Eirini %D 2024 %I SSRN %J Automation in Construction %K myown from:entoutsi aiml_group l3s %R 10.2139/ssrn.4703856 %T Generative AI based Augmentation for Offshore Jacket Design: An Integrated Approach for Mixed Tabular Data Generation under Data Scarcity and Imbalance %X Generative Artificial Intelligence (AI) has found various applications in domains like computer vision and natural language processing. However, limited research exists in the engineering domain, where prevailing challenges involve mixed tabular data, data scarcity, and imbalances. In this work, we focus on generating synthetic offshore jacket designs. Our goal is to improve the data quality of an existing scarce and imbalanced dataset. We quantify data quality by measuring the machine-learning efficiency of our synthetic data on a domain-specific downstream task. To this end, we propose an integrated method for generating jacket designs, combining modern data-driven and traditional multi-objective-driven approaches. Our method tackles mixed attributes, data scarcity, and imbalances. Experiments show improved predictive performance of models on the downstream task when trained on synthetic data vs solely real data. Our findings contribute to the advancement of generative AI in offshore engineering and related fields, offering valuable insights and potential applications. Keywords: Artificial Intelligence, Generative AI, Machine Learning, Multi-objective optimization, Mixed Tabular Data, Data Scarcity, Data Imbalance, Offshore Wind Turbines, Industrial Design, Offshore Jacket Substructure

@article{panagiotougenerative, abstract = {Generative Artificial Intelligence (AI) has found various applications in domains like computer vision and natural language processing. However, limited research exists in the engineering domain, where prevailing challenges involve mixed tabular data, data scarcity, and imbalances. In this work, we focus on generating synthetic offshore jacket designs. Our goal is to improve the data quality of an existing scarce and imbalanced dataset. We quantify data quality by measuring the machine-learning efficiency of our synthetic data on a domain-specific downstream task. To this end, we propose an integrated method for generating jacket designs, combining modern data-driven and traditional multi-objective-driven approaches. Our method tackles mixed attributes, data scarcity, and imbalances. Experiments show improved predictive performance of models on the downstream task when trained on synthetic data vs solely real data. Our findings contribute to the advancement of generative AI in offshore engineering and related fields, offering valuable insights and potential applications. Keywords: Artificial Intelligence, Generative AI, Machine Learning, Multi-objective optimization, Mixed Tabular Data, Data Scarcity, Data Imbalance, Offshore Wind Turbines, Industrial Design, Offshore Jacket Substructure}, added-at = {2024-02-28T20:34:02.000+0100}, author = {Panagiotou, Emmanouil and Qian, Han and Marx, Steffen and Ntoutsi, Eirini}, biburl = {https://www.bibsonomy.org/bibtex/2c0b3ed98e43291c6f01b0ef8d8dd126c/l3s}, doi = {10.2139/ssrn.4703856}, interhash = {8692d0e374fe47bab4704c927748b7ec}, intrahash = {c0b3ed98e43291c6f01b0ef8d8dd126c}, journal = {Automation in Construction}, keywords = {myown from:entoutsi aiml_group l3s}, language = {English}, location = {https://ssrn.com/paper=4703856}, publisher = {SSRN}, timestamp = {2024-02-28T20:34:02.000+0100}, title = {{Generative AI based Augmentation for Offshore Jacket Design: An Integrated Approach for Mixed Tabular Data Generation under Data Scarcity and Imbalance}}, type = {Working Paper Series}, year = 2024 }