MODEL LEARNING DEVICE, LABEL ESTIMATION DEVICE, METHOD THEREOF, AND PROGRAM

To provide a model learning device that performs model learning capable of performing label estimation with high accuracy even when learning data with few evaluators for individual data is used.SOLUTION: A model learning device performs learning processing using a set of a plurality of data and a la...

Full description

Saved in:
Bibliographic Details
Main Authors MASUMURA AKIRA, KOBASHIGAWA SATORU, KAMIYAMA HOSONA, ANDO ATSUSHI
Format Patent
LanguageEnglish
Japanese
Published 27.08.2020
Subjects
ACOUSTICS
CALCULATING
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
COMPUTING
COUNTING
MUSICAL INSTRUMENTS
PHYSICS
SPEECH ANALYSIS OR SYNTHESIS
SPEECH OR AUDIO CODING OR DECODING
SPEECH OR VOICE PROCESSING
SPEECH RECOGNITION
Online AccessGet full text

Cover

More Information
Summary:To provide a model learning device that performs model learning capable of performing label estimation with high accuracy even when learning data with few evaluators for individual data is used.SOLUTION: A model learning device performs learning processing using a set of a plurality of data and a label expected value, which is an index indicating degree of correctness of each label for the data, as learning data, and obtains a model for estimating a label for input data.SELECTED DRAWING: Figure 1 【課題】１データあたりの評価者が少ない学習データを用いた場合であっても、高い精度でラベル推定を行うことが可能なモデルを学習する。【解決手段】複数のデータと当該データに対する各ラベルの正しさの度合いを表す指標であるラベル期待値との組を学習データとして用いた学習処理を行い、入力データに対するラベルを推定するモデルを得る。【選択図】図１
Bibliography:Application Number: JP20190022353