Quantized kernel least mean square(QKLMS) algorithm is an effective nonlinear adaptive online learning algorithm with good performance in constraining the growth of network size through the use of quantization for inp...Quantized kernel least mean square(QKLMS) algorithm is an effective nonlinear adaptive online learning algorithm with good performance in constraining the growth of network size through the use of quantization for input space. It can serve as a powerful tool to perform complex computing for network service and application. With the purpose of compressing the input to further improve learning performance, this article proposes a novel QKLMS with entropy-guided learning, called EQ-KLMS. Under the consecutive square entropy learning framework, the basic idea of entropy-guided learning technique is to measure the uncertainty of the input vectors used for QKLMS, and delete those data with larger uncertainty, which are insignificant or easy to cause learning errors. Then, the dataset is compressed. Consequently, by using square entropy, the learning performance of proposed EQ-KLMS is improved with high precision and low computational cost. The proposed EQ-KLMS is validated using a weather-related dataset, and the results demonstrate the desirable performance of our scheme.展开更多
This paper proposed an universal steganalysis program based on quantification attack which can detect several kinds of data hiding algorithms for grayscale images. In practice, most techniques produce stego images tha...This paper proposed an universal steganalysis program based on quantification attack which can detect several kinds of data hiding algorithms for grayscale images. In practice, most techniques produce stego images that are perceptually identical to the cover images but exhibit statistical irregularities that distinguish them from cover images. Attacking the suspicious images using the quantization method, we can obtain statistically different from embedded-and-quantization attacked images and from quantization attacked-but-not-embedded sources. We have developed a technique based on one-class SVM for discriminating between cover-images and stego-images. Simulation results show our approach is able to distinguish between cover and stego images with reasonable accuracy.展开更多
A model for liquid-gas flow （MLGF）, considering the flee movement of liquid surface, was built to simulate the wastewater velocity field and gas distribution in a full-scale Caroussel oxidation ditch with surface ae...A model for liquid-gas flow （MLGF）, considering the flee movement of liquid surface, was built to simulate the wastewater velocity field and gas distribution in a full-scale Caroussel oxidation ditch with surface aeration. It was calibrated and validated by field measurement data, and the calibrated parameters and sections were selected based on both model analysis and numerical computation. The simulated velocities of MLGF were compared to that of a model for wastewater-sludge flow （MWSF）. The results show that the free liquid surface considered in MLGF improves the simulated velocity results of upper layer and surface. Moreover, distribution of gas volume fraction （GVF） simulated by MLGF was compared to dissolved oxygen （DO） measured in the oxidation ditch. It is shown that DO distribution is affected by many factors besides GVF distribution.展开更多
Objective To establish early detection and diagnosis for bladder cancer.Methods In the current study,a metabolomics strategy was used to profile bladder cancer urine metabolites in mice and to further characterize the...Objective To establish early detection and diagnosis for bladder cancer.Methods In the current study,a metabolomics strategy was used to profile bladder cancer urine metabolites in mice and to further characterize the disease status at different stages.In addition,some chemometrics algorithms were adopted to analyze the metabolites fingerprints,including baseline removal and retention time shift,to overcome variations in the experimental process.After processing,metabolites were qualitatively and quantitatively analyzed in each sample at different stages.Finally,a random forest algorithm was used to discriminate the differences among different groups.Results Four potential biomarkers,including glyceric acid,(R*,R*)-2,3-Dihydroxybutanoic acid,N-(1-oxohexyl)-glycine and D-Turanose,were discovered by exploring the characteristics of different groups.Conclusion These results suggest that combining chemometrics with the metabolites profile is an effective approach to aid in clinical diagnosis.展开更多
基金supported by the National Key Technologies R&D Program of China under Grant No. 2015BAK38B01the National Natural Science Foundation of China under Grant Nos. 61174103 and 61603032+4 种基金the National Key Research and Development Program of China under Grant Nos. 2016YFB0700502, 2016YFB1001404, and 2017YFB0702300the China Postdoctoral Science Foundation under Grant No. 2016M590048the Fundamental Research Funds for the Central Universities under Grant No. 06500025the University of Science and Technology Beijing - Taipei University of Technology Joint Research Program under Grant No. TW201610the Foundation from the Taipei University of Technology of Taiwan under Grant No. NTUT-USTB-105-4
文摘Quantized kernel least mean square(QKLMS) algorithm is an effective nonlinear adaptive online learning algorithm with good performance in constraining the growth of network size through the use of quantization for input space. It can serve as a powerful tool to perform complex computing for network service and application. With the purpose of compressing the input to further improve learning performance, this article proposes a novel QKLMS with entropy-guided learning, called EQ-KLMS. Under the consecutive square entropy learning framework, the basic idea of entropy-guided learning technique is to measure the uncertainty of the input vectors used for QKLMS, and delete those data with larger uncertainty, which are insignificant or easy to cause learning errors. Then, the dataset is compressed. Consequently, by using square entropy, the learning performance of proposed EQ-KLMS is improved with high precision and low computational cost. The proposed EQ-KLMS is validated using a weather-related dataset, and the results demonstrate the desirable performance of our scheme.
基金Science Fund of Shanghai Municipal Education Commission (03DZ13)
文摘This paper proposed an universal steganalysis program based on quantification attack which can detect several kinds of data hiding algorithms for grayscale images. In practice, most techniques produce stego images that are perceptually identical to the cover images but exhibit statistical irregularities that distinguish them from cover images. Attacking the suspicious images using the quantization method, we can obtain statistically different from embedded-and-quantization attacked images and from quantization attacked-but-not-embedded sources. We have developed a technique based on one-class SVM for discriminating between cover-images and stego-images. Simulation results show our approach is able to distinguish between cover and stego images with reasonable accuracy.
基金Project supported by Visiting Scholar Foundation of Key Laboratory of the Resources Exploitation and Environmental Disaster Control Engineering in Southwest China (Chongqing University),Ministry of Education,China
文摘A model for liquid-gas flow （MLGF）, considering the flee movement of liquid surface, was built to simulate the wastewater velocity field and gas distribution in a full-scale Caroussel oxidation ditch with surface aeration. It was calibrated and validated by field measurement data, and the calibrated parameters and sections were selected based on both model analysis and numerical computation. The simulated velocities of MLGF were compared to that of a model for wastewater-sludge flow （MWSF）. The results show that the free liquid surface considered in MLGF improves the simulated velocity results of upper layer and surface. Moreover, distribution of gas volume fraction （GVF） simulated by MLGF was compared to dissolved oxygen （DO） measured in the oxidation ditch. It is shown that DO distribution is affected by many factors besides GVF distribution.
基金funding support from the Natural Science Foundation of China (No. 81673585 and No. 81603400)Hunan Provincial Key Laboratory of Diagnostics in Chinese Medicine Open Fund (No. 2015ZYZD13 and No. 2015ZYZD10)+2 种基金Key research and development project of Hunan Province Science and Technology (No. 2016SK2048)Innovative Project for Post-graduate of Hunan University of Chinese Medicine (No. 2017CX05)the National Standard Project of Chinese Medicine (No. ZYBZH-Y-HUN-21)
文摘Objective To establish early detection and diagnosis for bladder cancer.Methods In the current study,a metabolomics strategy was used to profile bladder cancer urine metabolites in mice and to further characterize the disease status at different stages.In addition,some chemometrics algorithms were adopted to analyze the metabolites fingerprints,including baseline removal and retention time shift,to overcome variations in the experimental process.After processing,metabolites were qualitatively and quantitatively analyzed in each sample at different stages.Finally,a random forest algorithm was used to discriminate the differences among different groups.Results Four potential biomarkers,including glyceric acid,(R*,R*)-2,3-Dihydroxybutanoic acid,N-(1-oxohexyl)-glycine and D-Turanose,were discovered by exploring the characteristics of different groups.Conclusion These results suggest that combining chemometrics with the metabolites profile is an effective approach to aid in clinical diagnosis.
