Browsing by Author "Chiu, Sheng-Hsiung"

Now showing 1 - 5 of 5

A Hybrid Detecting Fraudulent Financial Statements Model Using Rough Set Theory and Support Vector Machines
(2016)
Yeh, Ching-Chiang
;
Chi, Der-Jang
;
Lin, Tzu-Yu
;
Chiu, Sheng-Hsiung
;
管理科學系
The detection of fraudulent financial statements (FFS) is an important and challenging issue that has served as the impetus for many academic studies over the past three decades. Although nonfinancial ratios are generally acknowledged as the key factor contributing to the FFS of a corporation, they are usually excluded from early detection models. The objective of this study is to increase the accuracy of FFS detection by integrating the rough set theory (RST) and support vector machines (SVM) approaches, while adopting both financial and nonfinancial ratios as predictive variables. The results showed that the proposed hybrid approach (RST+SVM) has the best classification rate as well as the lowest occurrence of Types I and II errors, and that nonfinancial ratios are indeed valuable information in FFS detection.
CLINICAL GENOMICS IDENTIFIES THE EXPRESSION OF HUMAN OVARIAN CARCINOMA
(2009-09-01)
Tsai, Meng-Hsiun
;
Huang, Steve G.
;
Tsaur, Shyh-Chang
;
Hung, Yao-Ching
;
Chiu, Sheng-Hsiung
Background: Human ovarian carcinoma is a silent killer. Most patients with human ovarian carcinoma are diagnosed in the advanced stages. Currently, no effective means of early detection are available. identifying genes that distinguish early cancers from normal tissue may provide insight to the initiation and progression of ovarian cancer. Methods: A large-scale gene expression profiling using 9,600 genes-complementary DNA array to study ovarian tumors was conducted. We analyzed 13 benign tumors, 22 invasive cancers (seven stage I and 15 stage III), and 16 normal ovaries. Cluster analysis revealed two major groups: the early-stage cluster that included 13 benign tumors and seven stage I cancers; and the late-stage cluster that included all stage III cancers. Results: There were 433 upregulated and 60 downregulated genes from the early-stage cluster, and 395 upregulated and 101 downregulated genes from the late-stage cluster. These genes showed altered functions for tumor/cancers, functions such as cell growth, cell cycle control, differentiation, signaling, apoptosis, ion transport, and morphogenesis. of most interest is the downregulation of nuclear receptor proteins. Differentially expressed genes between early-stage cluster and late-stage clusters were identified. We found 90 genes preferentially overly expressed and 165 genes relatively downregulated at late-stage cancers. The advanced cancer appears differently from early-stage cancer in cell growth, cell adhesion, hormone activity, morphology, and cell-cell signaling compared with benign or early-stage cancer. of particular interest are enzymes corresponding to pteridine biosynthesis, reported for the first time in this study. Real-time reverse transcription polymerase chain reaction showed that GCH1, KRT18, H3F3A, and MYBL2 were overly expressed in all cancer tissues, while MDFI, ARP3BETA, COL16A1, and MLN discriminated late-stage from early-stage cancers. Conclusion: The microarray provides valuable resources for comparing expression profiles of cancers at various stages and helps research the diagnosis and mechanism of ovarian cancer. [International journal of Gerontology 2009; 3(3): 163-169]
Corporate Performance Forecasting Using Hybrid Rough Set Theory, Neural Networks, and DEA
(2013-05-01)
Lin, Chiun-Sin
;
Lin, Tzu-Yu
;
Chiu, Sheng-Hsiung
;
管理科學系
;
Department of Management Science
This paper proposed the hybrid model using rough set theory (RST), neural networks (NN), and data envelopment analysis (DEA) to predict the corporate performance directly. First, to evaluate corporate performance, the DEA was employed. Second, integrated RST with BPN techniques, which is one of the popular used models of NN, named RST+BPN, was used to build the corporate performance-prediction model and the corporate governance variables are used as predictive variables. This hybrid method enabled us to evaluate an individual firm and provided performance information without comparing it with other companies. The experimental result showed that the proposed model outperforms the NN model with nonextracted predictive variables and provides a promising alternative in corporate performance prediction.
Empirical mode decomposition-based least squares support vector regression for foreign exchange rate forecasting
(2012-11-01)
Lin, Chiun-Sin
;
Chiu, Sheng-Hsiung
;
Lin, Tzu-Yu
;
管理科學系
;
Department of Management Science
To address the nonlinear and non-stationary characteristics of financial time series such as foreign exchange rates, this study proposes a hybrid forecasting model using empirical mode decomposition (EMD) and least squares support vector regression (LSSVR) for foreign exchange rate forecasting. EMD is used to decompose the dynamics of foreign exchange rate into several intrinsic mode function (IMF) components and one residual component. LSSVR is constructed to forecast these IMFs and residual value individually, and then all these forecasted values are aggregated to produce the final forecasted value for foreign exchange rates. Empirical results show that the proposed EMD-LSSVR model outperforms the EMD-ARIMA (autoregressive integrated moving average) as well as the LSSVR and ARIMA models without time series decomposition. (C) 2012 Elsevier B.V. All rights reserved.
Using independent component analysis and network DEA to improve bank performance evaluation
(2013-05-01)
Lin, Tzu-Yu
;
Chiu, Sheng-Hsiung
;
管理科學系
;
Department of Management Science
Given the importance of accurate performance evaluation for the banking industry, this paper uses data on Taiwanese domestic banks and investigates whether it is useful to construct an integrated performance model in order to address the problems of efficiency-analysis with the aid of an independent component analysis (ICA) and a network slacks-based measure (NSBM). ICA is related to the search for latent information: in it, independent components (ICs) from the observed data and selected statistically unrelated ICs are used as new input, intermediate, and output variables for the NSBM. The NSBM is then used in the investigation of multiple-dimension efficiencies, along with operating performance, in the Taiwanese domestic banking sector. The results show that the proposed ICA-NSBM model provides sufficient information to determine the main sources of inefficiency at the dimensional level and demonstrates excellent significant discriminative capability. (C) 2013 Elsevier B.V. All rights reserved.