The financial crisis prediction model and critical success factors analysis of the real estate development company using data mining
| Autor: | Li-Wei Huang, 黃立維 |
|---|---|
| Rok vydání: | 2008 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 96 Investment in the construction industry, the study referred to as the construction industry, its economy and has close relations, therefore, the construction industry Construction of a financial crisis prediction model for the relevant decision makers, essential. However, the construction industry for the past financial crisis model of research, focusing on financial information and with statistical model, this study will combine data mining and general economic variables as the basis for modeling. This study was to take samples in 88-91 years a total of 23 samples of the crisis the company to match 1:1 normal mode selected 23 companies, a total of 46 samples of the company to financial information and the overall rate of economic and business cycle, and other variables as input The indicators. This study through to data mining in the neural networks and decision trees algorithms construct a financial crisis prediction model, the results show that neural network performance is better than the forecast decision tree model, the neural network were to construct a financial The average pre-crisis three-year-period data from the "average earnings prediction mode" performance better than forecast, "the report prediction mode", which again did not join the general economic variables A-4 model of best 91.6 percent, and by factors of the financial screening Before the one-year-period crisis, before two-year-period three years before the common factors, the financial crisis that affected the key variables are: debt ratio, quick ratio, debt/eqity, the inventory turnover rate, earnings per share by the decision tree The tree was informed that the rules of the financial crisis prediction capacity of more difference between a variable rate liabilities, respectively, debt ratio, earnings per share, the debt/eqity and overall economic variables and can not effectively improve forecast accuracy rate, after the last test case of the test group, Further proof of the reliability model, for the reference of the decision-making. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|