Traditional data driven fault detection methods assume that the process operates in a single mode so that they cannot perform well in processes with multiple operating modes. To monitor multimode processes effectively...Traditional data driven fault detection methods assume that the process operates in a single mode so that they cannot perform well in processes with multiple operating modes. To monitor multimode processes effectively,this paper proposes a novel process monitoring scheme based on orthogonal nonnegative matrix factorization(ONMF) and hidden Markov model(HMM). The new clustering technique ONMF is employed to separate data from different process modes. The multiple HMMs for various operating modes lead to higher modeling accuracy.The proposed approach does not presume the distribution of data in each mode because the process uncertainty and dynamics can be well interpreted through the hidden Markov estimation. The HMM-based monitoring indication named negative log likelihood probability is utilized for fault detection. In order to assess the proposed monitoring strategy, a numerical example and the Tennessee Eastman process are used. The results demonstrate that this method provides efficient fault detection performance.展开更多
The data envelopment analysis (DEA) model is used to evaluate the relative economic efficiency of a given set of decision making units (DMUs). In this paper, the DEA production possibility set is transferred from ...The data envelopment analysis (DEA) model is used to evaluate the relative economic efficiency of a given set of decision making units (DMUs). In this paper, the DEA production possibility set is transferred from the conventional sum form into the intersection form which is represented by a linear inequality system. Although it is time consuming to obtain the intersection form of the production possibility set, it suggests a new angle to investigate the properties of DMUs and to extend the DEA research further beyond the efficiency measurement. Following the intersection form, the analytical formula of the efficiency indicator and projection is given. Various aspects of technical efficiency, returns to scale and evidence of congestion of the DMUs are studied. The relationship between the weak DEA efficiency and the weak Pareto solution is discussed. Finally, a procedure for DMU grouping is proposed to help the decision makers for better resource reallocation and strategy adjustment.展开更多
The paper proposes a new approach -- The decomposition-based vector autoregressive (DVAR) model to scrutinize the predictability of the UK stock market. Empirical studies performed on the monthly British FTSE100 ind...The paper proposes a new approach -- The decomposition-based vector autoregressive (DVAR) model to scrutinize the predictability of the UK stock market. Empirical studies performed on the monthly British FTSE100 index over 1984-2012 confirm that the DVAR model does provide informative forecasts for both in-sample and out-of-sample forecasts. Trading strategies based on the DVAR forecasts can Significantly beat the simple buy-and-hold, which demonstrates the valuable information provided by technical analysis in the UK stock market.展开更多
The authors give a discription of the finite representation type over an algebraically stable categories of selfinjective algebras of closed field, which admits indecomposable Calabi-Yau obdjects. For selfinjective al...The authors give a discription of the finite representation type over an algebraically stable categories of selfinjective algebras of closed field, which admits indecomposable Calabi-Yau obdjects. For selfinjective algebras with such properties, the ones whose stable categories are not Calabi-Yau are determined. For the remaining ones, i.e., those selfinjective algebras whose stable categories are actually Calabi-Yau, the difference between the Calabi-Yau dimensions of the indecomposable Calabi-Yau objects and the Calabi-Yau dimensions of the stable categories is described.展开更多
We present a novel model for recognizing long-term complex activities involving multiple persons. The proposed model, named ‘decomposed hidden Markov model’ (DHMM), combines spatial decomposition and hierarchical ab...We present a novel model for recognizing long-term complex activities involving multiple persons. The proposed model, named ‘decomposed hidden Markov model’ (DHMM), combines spatial decomposition and hierarchical abstraction to capture multi-modal, long-term dependent and multi-scale characteristics of activities. Decomposition in space and time offers conceptual advantages of compaction and clarity, and greatly reduces the size of state space as well as the number of parameters. DHMMs are efficient even when the number of persons is variable. We also introduce an efficient approximation algorithm for inference and parameter estimation. Experiments on multi-person activities and multi-modal individual activities demonstrate that DHMMs are more efficient and reliable than familiar models, such as coupled HMMs, hierarchical HMMs, and multi-observation HMMs.展开更多
The authors prove that an operator with the cellular indecomposable property has no singular points in the semi-Fredholm domain, by applying the 4 x 4 matrix model of semi-Fredholm operators due to Fang in 2004. This ...The authors prove that an operator with the cellular indecomposable property has no singular points in the semi-Fredholm domain, by applying the 4 x 4 matrix model of semi-Fredholm operators due to Fang in 2004. This result fills a gap in the result of Olin and Thomson in 1984.展开更多
基金Supported by the National Natural Science Foundation of China(61374140,61403072)
文摘Traditional data driven fault detection methods assume that the process operates in a single mode so that they cannot perform well in processes with multiple operating modes. To monitor multimode processes effectively,this paper proposes a novel process monitoring scheme based on orthogonal nonnegative matrix factorization(ONMF) and hidden Markov model(HMM). The new clustering technique ONMF is employed to separate data from different process modes. The multiple HMMs for various operating modes lead to higher modeling accuracy.The proposed approach does not presume the distribution of data in each mode because the process uncertainty and dynamics can be well interpreted through the hidden Markov estimation. The HMM-based monitoring indication named negative log likelihood probability is utilized for fault detection. In order to assess the proposed monitoring strategy, a numerical example and the Tennessee Eastman process are used. The results demonstrate that this method provides efficient fault detection performance.
基金This research is supported by the National Natural Science Foundation of China under Grant Nos. 70531040, 70871114, and the 985 Research Grant of Renmin University of China, and the Hong Kong CERG Research Fund PolyU5457/06H and PolyU 5485/09H.
文摘The data envelopment analysis (DEA) model is used to evaluate the relative economic efficiency of a given set of decision making units (DMUs). In this paper, the DEA production possibility set is transferred from the conventional sum form into the intersection form which is represented by a linear inequality system. Although it is time consuming to obtain the intersection form of the production possibility set, it suggests a new angle to investigate the properties of DMUs and to extend the DEA research further beyond the efficiency measurement. Following the intersection form, the analytical formula of the efficiency indicator and projection is given. Various aspects of technical efficiency, returns to scale and evidence of congestion of the DMUs are studied. The relationship between the weak DEA efficiency and the weak Pareto solution is discussed. Finally, a procedure for DMU grouping is proposed to help the decision makers for better resource reallocation and strategy adjustment.
基金supported by Social Science Foundation of Ministry of Education of China under Grant No.12YJC790001National Social Science Foundation of China under Grant No.12CJY117+1 种基金the National Natural Science Foundation of China under Grant Nos.71003057 and 71373262the Program for Innovative Research Team and“211”Program in UIBE
文摘The paper proposes a new approach -- The decomposition-based vector autoregressive (DVAR) model to scrutinize the predictability of the UK stock market. Empirical studies performed on the monthly British FTSE100 index over 1984-2012 confirm that the DVAR model does provide informative forecasts for both in-sample and out-of-sample forecasts. Trading strategies based on the DVAR forecasts can Significantly beat the simple buy-and-hold, which demonstrates the valuable information provided by technical analysis in the UK stock market.
基金supported by the National Natural Science Foundation of China (No. 10801099)the Zhejiang Provincial Natural Science Foundation of China (No. J20080154)the grant from Science Technology Department of Zhejiang Province (No. 2011R10051)
文摘The authors give a discription of the finite representation type over an algebraically stable categories of selfinjective algebras of closed field, which admits indecomposable Calabi-Yau obdjects. For selfinjective algebras with such properties, the ones whose stable categories are not Calabi-Yau are determined. For the remaining ones, i.e., those selfinjective algebras whose stable categories are actually Calabi-Yau, the difference between the Calabi-Yau dimensions of the indecomposable Calabi-Yau objects and the Calabi-Yau dimensions of the stable categories is described.
基金Project (No. 60772050) supported by the National Natural Science Foundation of China
文摘We present a novel model for recognizing long-term complex activities involving multiple persons. The proposed model, named ‘decomposed hidden Markov model’ (DHMM), combines spatial decomposition and hierarchical abstraction to capture multi-modal, long-term dependent and multi-scale characteristics of activities. Decomposition in space and time offers conceptual advantages of compaction and clarity, and greatly reduces the size of state space as well as the number of parameters. DHMMs are efficient even when the number of persons is variable. We also introduce an efficient approximation algorithm for inference and parameter estimation. Experiments on multi-person activities and multi-modal individual activities demonstrate that DHMMs are more efficient and reliable than familiar models, such as coupled HMMs, hierarchical HMMs, and multi-observation HMMs.
基金Project supported by the National Natural Science Foundation of China(No.11101312)the National Science Foundation(No.0801174)
文摘The authors prove that an operator with the cellular indecomposable property has no singular points in the semi-Fredholm domain, by applying the 4 x 4 matrix model of semi-Fredholm operators due to Fang in 2004. This result fills a gap in the result of Olin and Thomson in 1984.