一個有效的繼續經營預測模型

CCUR > College of Business > Department of Accounting & Graduate Institute of Accounting > Thesis > Item 987654321/48496

Please use this identifier to cite or link to this item: https://irlib.pccu.edu.tw/handle/987654321/48496

Title:	一個有效的繼續經營預測模型 An Effective Model of Going Concern Detection
Authors:	黃奕文
Contributors:	會計學系
Keywords:	繼續經營預測類神經網路逐步迴歸決策樹區別分析
Date:	2020
Issue Date:	2020-08-26 14:04:29 (UTC+8)
Abstract:	本研究主要目的是應用機器學習的技術來建構繼續經營預測模型，首先運用類神經網路篩選出重要變數，再配合C5.0決策樹模式來建立預測模型。為了測試本研究所提模型的有效性，本研究資料來自台灣經濟新報社資料庫中(TEJ)，研究對象為2009年至2018年財務報表有繼續經營疑慮及無繼續經營疑慮之上市櫃公司(以1:3進行配對)，研究變數涵蓋財務與非財務變數。相較於傳統的統計模式對繼續經營意見的判斷，以機器學習技術建立的預測模式有較優秀的預測績效，其中ANN搭配C5.0建構的繼續經營預測模型整體預測準確率達92.49%。 The main purpose of this study is to apply the machine learning techniques to construct a prediction model for going concern. First, the artificial neural network is used to screen out important variables, and then the C5.0 decision tree model is used to establish the prediction model. The data of this study is from the Taiwan Economic Journal (TEJ) from 2009 to 2018. The research object is a listed company that has doubts about going concern and no doubts about going concern (1GC sample: 3Non-GC samples). The research variables cover financial and non-financial variables. Compared with the traditional statistical model for the judgment of going concern opinions, the prediction model established by machine learning technology has better prediction performance. The overall prediction accuracy rate of ANN-C5.0 model is 92.49%.
Appears in Collections:	[Department of Accounting & Graduate Institute of Accounting] Thesis

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	175	View/Open

View Licence

Loading...