US 12,354,018 B2
Prediction management system, prediction management method, data structure, prediction management device and prediction execution device
Takuya Kadohira, Tsukuba (JP); Satoshi Minamoto, Tsukuba (JP); Makoto Watanabe, Tsukuba (JP); Masahiko Demura, Tsukuba (JP); Junya Inoue, Tokyo (JP); and Manabu Enoki, Tokyo (JP)
Assigned to NATIONAL INSTITUTE FOR MATERIALS SCIENCE, Tsukuba (JP)
Appl. No. 17/279,680
Filed by NATIONAL INSTITUTE FOR MATERIALS SCIENCE, Tsukuba (JP)
PCT Filed Oct. 1, 2019, PCT No. PCT/JP2019/038709
§ 371(c)(1), (2) Date Mar. 25, 2021,
PCT Pub. No. WO2020/075573, PCT Pub. Date Apr. 16, 2020.
Claims priority of application No. 2018-191904 (JP), filed on Oct. 10, 2018.
Prior Publication US 2021/0397976 A1, Dec. 23, 2021
Int. Cl. G06N 5/02 (2023.01)
CPC G06N 5/02 (2013.01) 10 Claims
OG exemplary drawing
 
1. A prediction management system that makes a prediction regarding a material, the prediction management system comprising:
a processor; and
a non-transitory memory storing instructions that, when executed by the processor, cause the processor to perform operations comprising:
storing descriptors in a descriptor storage, each descriptor of the descriptors describing a parameter regarding processing means, a structure, a property, or a performance of the material;
storing prediction models in a prediction model storage, each prediction model of the prediction models describing an input and output relationship among the descriptors, wherein each prediction model accepts, as an input descriptor, one of at least two of the processing means, the structure, the property, and the performance, and outputs another one thereof as an output descriptor;
storing workflows in a workflow storage, each workflow of the workflows describing that at least two of the prediction models are connected to each other via a descriptor, wherein the output descriptor of one prediction model is accepted as the input descriptor of another prediction model;
giving an input to each workflow of the workflows, executing the workflow including execution of each prediction model included in the workflow, and storing an execution result including an output result of each prediction model included in the workflow indicating the prediction regarding the material; and
managing execution results, the workflows, the prediction models, and the descriptors in four hierarchical layers by assigning a unique identifier.