Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
機械学習システムのトレーニング方法及びトレーニングシステム
Document Type and Number:
Japanese Patent JP7004661
Kind Code:
B2
Abstract:
A training method and a training system for a machine learning system are disclosed in the present application. Training data is used for training the machine learning system. The training method includes allocating the training data to a plurality of working machines; dividing training data allocated by each working machine into a plurality of data pieces; obtaining a local weight and a local loss function value calculated by each working machine based on each data piece; aggregating the local weight and the local loss function value calculated by each work machine based on each data piece to obtain a current weight and a current loss function value; performing model abnormality detection using the current weight and/or the current loss function value; inputting a weight and a loss function value of a previous aggregation to the machine learning system for training in response to a result of the model abnormality detection being a first type of abnormality; and modifying the current weight and/or the current loss function value to a current weight and/or a current loss function value within a first threshold in response to the result of the model abnormality detection being a second type of abnormality, and inputting thereof to the machine learning system for training.

Inventors:
Jou Jun
Application Number:
JP2018546445A
Publication Date:
January 21, 2022
Filing Date:
February 21, 2017
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
ALIBABA GROUP HOLDING LIMITED
International Classes:
G06N20/00
Domestic Patent References:
JP2009288933A
JP2012022558A
JP2001344590A
Other References:
浦 晃 ほか,「分散計算環境における並列パーセプトロンの将棋評価関数への適用」,情報処理学会研究報告 ゲーム情報学(GI) [online],情報処理学会,2014年03月10日,第2014-GI-31巻, 第6号,pp.1-8,[2014年3月11日検索],インターネット:
Attorney, Agent or Firm:
Patent Business Corporation Tani / Abe Patent Office