Applying Machine Learning Technique to Construct the Structure of the Board of Directors and Company Performance Prediction Model

• Main Authors: Nieh,Yu-Hua, 聶宇華
• Other Authors: Huang,Mu-Jung
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/d3fa9v

碩士 === 國立彰化師範大學 === 會計學系 === 107 === Because of the conflict in international trade, Taiwan's traditional industrial companies need to rethink the direction of future development through the model of taking China as their production base. Therefore, how the board of directors will lead to the sustainable development of traditional industrial companies in the future situation will have a significant impact on the development of the nation and the growth of enterprises. During the research period from 2008 to 2017, filtered traditional industrial companies to conducted research, using the C4.5 algorithm to obtain the maximum amount of information gain, and utilize PEP pruning to avoid the estimate of over optimistic. In addition, continuous correction of the binomial distributions is also be carried out. In order to visualize the Apriori algorithm, the clustering algorithm of K-means is be integrated so that can effectively reduce the number of rules and the problem which is difficult to analyze. Based on the theory of agency, this research makes use of the C4.5 algorithm to consider the relationship between the board of directors, the manager, the institutional investor and the corporate performance of the corporate governance. Using Apriori algorithm to clarified and link the factor association between MSOWN, BIG4, MDUAL, which are relatively low in decision tree probability but relatively strong to the company's performance.