ACTION LEARNING DEVICE, ACTION LEARNING METHOD, ACTION DETERMINATION DEVICE, ACTION DETERMINATION METHOD, ACTION LEARNING SYSTEM, PROGRAM, AND RECORDING MEDIUM

Title:

ACTION LEARNING DEVICE, ACTION LEARNING METHOD, ACTION DETERMINATION DEVICE, ACTION DETERMINATION METHOD, ACTION LEARNING SYSTEM, PROGRAM, AND RECORDING MEDIUM

Document Type and Number:

WIPO Patent Application WO/2021/025094

Kind Code:

A1

Abstract:

An action learning device comprising: an action select unit for selecting, on the basis of state information that represents an environment and a self-state, an action candidate that is executed with respect to the environment; an evaluation acquisition unit for acquiring the evaluation of a user with respect to the action candidate selected by the action select unit, the evaluation indicating, together with a reason, the determination to execute or not execute the action candidate in the state indicated by the state information data; a slot generation unit for generating, on the basis of the reason in the evaluation, a slot that indicates a location of interest in the state information data; and a user learning model generation unit for generating a user learning model in which the state information data, the slot, and the determination in the evaluation are linked to the action candidate.

More Like This:

JPH0784977	METHOD FOR CONSTITUTING MULTILAYER STRUCTURE TYPE NEURAL NETWORK
WO/2022/055828	A MEMORY INCLUDING EXAMPLES OF CALCULATING HAMMING DISTANCES FOR NEURAL NETWORK AND DATA CENTER APPLICATIONS
WO/2022/129156	EXPLOITATION OF LOW DATA DENSITY OR NONZERO WEIGHTS IN A WEIGHTED SUM COMPUTER

Inventors:

MIYAUCHI YOSHIHITO (JP)
UDA AKIO (JP)
YAMAMOTO KYOUSEI (JP)

Application Number:

PCT/JP2020/030111

Publication Date:

February 11, 2021

Filing Date:

August 06, 2020

Export Citation:

Click for automatic bibliography generation Help

Assignee:

NEC SOLUTION INNOVATORS LTD (JP)

International Classes:

G06N3/04; G06N3/08

Domestic Patent References:

WO2019022085A1

2019-01-31

Foreign References:

CN109978012A

2019-07-05

Other References:

MATSUI, KAZUAKI; MATOBA, RYUICHI: "Selections of Discarding Mahjong Piece Using Neural Network", IPSJ SIG TECHNICAL REPORTS: GAME INFORMATICS, vol. 2015-GI-34, no. 8, 27 June 2015 (2015-06-27), pages 1 - 5, XP009526806
RUPENEITE ANNIJA: "Building Poker Agent Using Reinforcement Learning with Neural Networks", SCITEPRESS DIGITAL LIBRARY, 1 January 2014 (2014-01-01), XP055792164, Retrieved from the Internet [retrieved on 20200907]

Attorney, Agent or Firm:

OKABE, Yuzuru et al. (JP)

Download PDF:

View/Download PDF PDF Help

Previous Patent: MOISTURE-CURABLE POLYURETHANE HOT-MELT RESIN COMPOSITION

Next Patent: SLURRY TRANSFER FACILITY AND SLURRY FEEDING METHOD