実質破綻の早期発見を目指したブースティングによる予測モデルの構築  [in Japanese] Early Detection Model for Business Failure Using AdaBoost  [in Japanese]

Access this Article

Author(s)

Abstract

多くの倒産予測研究では,財務指標選択と予測モデル構築が別々に行われており,両者を含めた処理全体としての予測精度の最適性が保証されていない.本研究では,実質破綻予測のための財務指標選択と予測モデル構築を一貫した枠組みで行うために,AdaBoostアルゴリズムを利用する.財務指標候補として,時系列の財務諸表に含まれる任意の二つの会計項目を比率の形で用意し,数年先に実質破綻に陥る企業の予測に有効な財務比率の抽出と識別関数の導出をAdaBoostにより行う.評価実験により,複数の会計年度にまたがる財務指標を用いることで,破綻の数年前の段階で実質破綻企業と継続企業を高い精度で識別できることが示された.

Critical elements in the process of predicting business failures are selection of financial indicators and construction of a forecast model. Many traditional methods implement these two tasks separately, and do not guarantee the optimality on the whole process. This study uses the AdaBoost algorithm so that these two processes can be realized within a single coherent framework. Setting financial ratios generated from arbitrary two items in the time-series financial data as candidates of indicators, our proposed method selects the effective financial ratios and derives the discrimination function which can predict companies going bankrupt in a few years. Experimental results show that our method can distinguish failed firms from continuing ones with higher accuracy by using the proposed temporal financial ratios.

Journal

  • Abstracts of Annual Conference of Japan Society for Management Information

    Abstracts of Annual Conference of Japan Society for Management Information 2017s(0), 44-47, 2017

    THE JAPAN SOCIETY FOR MANAGEMENT INFORMATION (JASMIN)

Codes

Page Top