IMPORTANCE ANALYSIS DEVICE, METHOD, AND PROGRAM

To improve analysis accuracy for importance of a feature in machine learning.SOLUTION: An importance analysis device has an importance calculation unit and a distribution calculation unit. The importance calculation unit calculates the importance of each of a plurality of features of each of a plura...

Full description

Saved in:
Bibliographic Details
Main Authors NAKATA KOTA, TAKAKI KENTARO
Format Patent
LanguageEnglish
Japanese
Published 23.03.2022
Subjects
CALCULATING
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
COMPUTING
COUNTING
IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
PHYSICS
Online AccessGet full text

Cover

More Information
Summary:To improve analysis accuracy for importance of a feature in machine learning.SOLUTION: An importance analysis device has an importance calculation unit and a distribution calculation unit. The importance calculation unit calculates the importance of each of a plurality of features of each of a plurality of input data based on a trained model and the plurality of input data. The distribution calculation unit calculates the distribution of the importance for each of the plurality of features over the plurality of input data.SELECTED DRAWING: Figure 2 【課題】機械学習における特徴量の重要度の解析精度を向上させること。【解決手段】 実施形態に係る重要度解析装置は、重要度算出部と分布算出部とを有する。重要度算出部は、学習済みモデルと複数個の入力データとに基づいて前記複数個の入力データ各々の複数の特徴量毎の重要度を算出する。分布算出部は、前記複数個の入力データに亘る前記複数の特徴量毎の前記重要度の分布を算出する。【選択図】 図２
Bibliography:Application Number: JP20200151951