MACHINE LEARNING SYSTEM AND MACHINE LEARNING METHOD
To provide a machine learning system capable of comprehending the influences on the performance of a machine learning model, which is rendered by exclusion or addition of evaluation target data to training data.SOLUTION: A machine learning system 100 includes: an acquisition unit that acquires an in...
Gespeichert in:
Hauptverfasser: | , , |
---|---|
Format: | Patent |
Sprache: | eng ; jpn |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | To provide a machine learning system capable of comprehending the influences on the performance of a machine learning model, which is rendered by exclusion or addition of evaluation target data to training data.SOLUTION: A machine learning system 100 includes: an acquisition unit that acquires an initial data group used for learning a learning model, an evaluation target data to be added or excluded from the initial data group, and a verification data group that includes at least one element that is not included in the evaluation target data; and a contribution calculation unit that calculates the contribution to the evaluation of the effect of the data to be evaluated on the performance of the learning model based on the output value by the training model that inputs the verification data group and the output value by the re-learning model that is trained by adding or excluding the evaluation target data to the initial data group.SELECTED DRAWING: Figure 1
【課題】学習データに対する評価対象データの除外及び追加が機械学習モデルの性能に与える影響の良否を知り得るようする。【解決手段】機械学習システム100は、学習モデルの学習に用いた初期データ群、初期データ群に対して追加又は除外する評価対象データ、及び評価対象データに含まれない要素を少なくとも1つ含む検証データ群を取得する取得部と、検証データ群を入力とする、学習モデルによる出力値と、初期データ群に対して評価対象データを追加又は除外して学習した再学習モデルによる出力値とに基づいて、評価対象データが学習モデルの性能に与える影響を評価する寄与度を計算する寄与度計算部と、を有する。【選択図】 図1 |
---|