仿生資料庫

    In the material manufacturing industry, whether it is formula design, process parameters or back-end product characteristics, most of them are conducted in the form of trial-and-error, which consumes a lot of manpower, resources and time. And the development cycle of new materials or new products often takes many years. Due to the significant advancement in computer computing performance and data analysis technology, from finance, e-commerce to the medical industry, it is an inevitable development direction to master key technologies or products through massive data analysis. In the field of materials, the combination with artificial intelligence machine learning technology has also become an international trend. 

    Machine learning uses mathematical algorithms to train the data to produce predictive models. Through this model, the appropriate parameters are selected before the experiment, which accelerates the time course of material development and lays out the next generation materials and products. In the material industry, data accumulation is not easy, so how to apply machine learning technology to small sample data is a technical bottleneck. The Material Genome Initiative (MGI) in the United States since 2011 invested a large amount of resources in the development of high-throughput calculations, high-throughput experiments, and collecting information on the structure and characteristics of materials as a digital database, through appropriate organization. It is used in material design and retrieval, and cross-domain integration technology with artificial intelligence machine learning, and a shared platform for the material innovation, which is the core content of the new material research methodology in the Information Age.