Search academic texts

Information × Registration Number 0225U004637, (0124U003279) , R & D reports Title Machine Learning Enhanced Additive Manufacturing popup.stage_title Розробка сурогатних (базових) моделей машинного навчання для методу Робокастинг Head Zhalat-Lozynskyi Ostap B., Доктор технічних наук Registration Date 15-12-2025 Organization Frantsevich Institute for Problems of Materials Science National Academy of Science of Ukraine popup.description1 The goal of the project is to develop a new, data-driven approach to the control of the additive manufacturing process through the application of innovative principles based on machine learning. popup.description2 Сonducted research on the development of surrogate (base) machine learning models that predict the values ​​of deviations of geometric dimensions of printed products of simple shape depending on the printing parameters, and are also capable of quantitatively assessing the model's uncertainty in the accuracy of the forecast for a specific combination of printing parameters. A number of machine learning models with different regression algorithms were built and their predictive ability was analyzed. According to the results, regression on Gaussian processes was selected as the main algorithm that satisfies the prediction conditions for both the prediction of deviation values ​​and the model's uncertainty regarding their forecast, which allow Bayesian optimization of the experiment and the model at the next stage of work. Product Description popup.authors OSTAP B. ZGHALAT-LOZYNSKYI Nazarii R. Mediukh Vasiliev Oleksandr O. Naumenko Vladyslav Yu. Krasikov Anton I. Shevchuk Kateryna S. Bilyi Vladyslav O. popup.nrat_date 2025-12-15 Close