INFORMATION PROCESSING DEVICE, INFORMATION PROCESSING METHOD AND INFORMATION PROCESSING PROGRAM

To simply correct training data collected from a parent population and having a bias and train a high-accuracy mathematical model by machine learning, using the training data after correction.SOLUTION: An information processing device according to the present invention comprises: a simultaneous dist...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: TAKEUCHI WATARU, NAKAGAWA SHUNKI
Format: Patent
Sprache:eng ; jpn
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:To simply correct training data collected from a parent population and having a bias and train a high-accuracy mathematical model by machine learning, using the training data after correction.SOLUTION: An information processing device according to the present invention comprises: a simultaneous distribution creation unit that creates, from training data, simultaneous distribution information having the same variables as statistical information; a weight calculation unit that calculates a bias degree, by which degree the sample of the created simultaneous distribution information is biased from the statistical information as a parent population, as a weight for each of a plurality of combinations of the variables; a prediction model creation unit that corrects the training data on the basis of the calculated weight, and trains a prediction model using the corrected training data; and an output processing unit that outputs a prediction result with respect to prediction data, using the prediction model.SELECTED DRAWING: Figure 1 【課題】母集団から収集した偏りのある学習用データを簡便に修正し、修正後の学習用データを使用して、精度の高い数理モデルを機械学習する。【解決手段】本発明の情報処理装置は、統計情報と同じ変数を有する同時分布情報を学習用データから作成する同時分布作成部と、前記作成した同時分布情報の標本が母集団としての前記統計情報に比して偏っている程度を、複数の前記変数の組み合わせごとに重みとして算出する重み算出部と、前記算出した重みに基づいて、前記学習用データを修正し、前記修正した学習用データを使用して予測モデルを学習する予測モデル作成部と、前記予測モデルを使用して予測用データに対する予測結果を出力する出力処理部と、を備えることを特徴とする。【選択図】図1