JMP/Python Ecosystem Success Stories (2023-KR-30MP-1556)Kwangki Lee, Haneol Solution
The JMP/Python ecosystem basically means an ecosystem that uses JMP's functions and Python's functions in a complementary way. Especially from a machine learning perspective, JMP Pro's own machine learning techniques as well as Python's own We first built a JMP/Python machine learning ecosystem that builds and optimizes machine learning prediction models using scikit-learn, XGBoost, LightGBM, CatBoost, PyTorch, and Tensor Flow in a complementary manner. The roadmap of the JMP/Python ecosystem is defined as IGEO, and the IGEO roadmap is roughly divided into the IG (Identify/Generate) area and the EO (Explore/Optimize) area. The IG area is the problem definition area and the area that generates experiment/simulation data. The EO area, the core of the JMP/Python ecosystem, is an area that executes DEbO (Design Exploration based Optimization). In the EO area, design is performed by integrating MariaDB, JMP, and Python, and if necessary, automation is performed by building a GUI using Python/JMP GUI. To explore the design area, the two steps of creating an experimental design/computerized experimental design and building a machine learning model are repeated, and JMP and Python modules are combined depending on the situation. As a success story of the JMP/Python machine learning ecosystem, we have summarized the Explore area cases, ball bearing failure detection monitoring system cases, and CAE machine learning optimization cases, and the combination of JMP and Python provides a direction for performing almost all engineering analysis. I would like to present it.
