Title:
INFERENCE DEVICE, INFERENCE METHOD, AND INFERENCE PROGRAM
Document Type and Number:
WIPO Patent Application WO/2021/014878
Kind Code:
A1
Abstract:
An inference device according to one aspect of the present invention supplies subject data to each of a plurality of inference models derived so as to carry out a prescribed inference from local learning data acquired under different environments, and causes each inference model to carry out a prescribed inference, thereby acquiring an inference result of each inference model. The inference device then determines a value of each integration parameter in accordance with environment data, uses the determined value of each integration parameter to assign a weight to the inference result of each inference model, and integrates the weighted inference results, thereby generating an inference result for the subject environment.
Inventors:
YONETANI RYO (JP)
SUWA MASAKI (JP)
BAREKATAIN MOHAMMADAMIN (JP)
IJIRI YOSHIHISA (JP)
MIYAURA HIROYUKI (JP)
SUWA MASAKI (JP)
BAREKATAIN MOHAMMADAMIN (JP)
IJIRI YOSHIHISA (JP)
MIYAURA HIROYUKI (JP)
Application Number:
PCT/JP2020/024998
Publication Date:
January 28, 2021
Filing Date:
June 25, 2020
Export Citation:
Assignee:
OMRON TATEISI ELECTRONICS CO (JP)
International Classes:
G06N20/00
Foreign References:
JP2019095968A | 2019-06-20 | |||
JP2018072876A | 2018-05-10 | |||
JP2019125113A | 2019-07-25 | |||
JP2012038794A | 2012-02-23 | |||
JP2012216158A | 2012-11-08 | |||
JP2011059924A | 2011-03-24 | |||
JP2017173874A | 2017-09-28 | |||
JP2018051664A | 2018-04-05 | |||
JP2019067238A | 2019-04-25 | |||
JP2019087096A | 2019-06-06 |
Other References:
NAOKI KOTANIKENJI TANIGUCHI: "Directivity Design of Ultrasonic Array Sensor Based on Reinforcement Learning", TRANSACTIONS OF THE INSTITUTE OF SYSTEMS, CONTROL AND INFORMATION ENGINEERS, vol. 23, no. 12, 2010, pages 291 - 293
JUSTIN FU ET AL.: "Learning Robust Rewards with Adversarial Inverse Reinforcement Learning", ARXIV:1710.11248, 2018
See also references of EP 4006790A4
JUSTIN FU ET AL.: "Learning Robust Rewards with Adversarial Inverse Reinforcement Learning", ARXIV:1710.11248, 2018
See also references of EP 4006790A4
Attorney, Agent or Firm:
TACHIBANA, Kenji et al. (JP)
Download PDF: