Time filter

Source Type

Shangzhou, China

Ning L.,Shangluo University
Naukovyi Visnyk Natsionalnoho Hirnychoho Universytetu | Year: 2016

Purpose: In practical application, the accuracy of the minority class is very important and the research on imbalanced data has become one of the most popular topics. In order to improve the classification performance for imbalanced data, the classification algorithm based on data sampling and integration technology for imbalanced data was proposed. Methodology: Firstly, the traditional SMOTE algorithm was improved to K-SMOTE (an over-sampling method based on SMOTE and K-means). In K-SMOTE, the dataset was to perform clustering operation, and the interpolation operation was performed on the connection of the cluster center and the original data point. Secondly, ECA-IBD (an ensemble classification algorithm based improved SMOTE for imbalanced data) was proposed. In ECA-IBD, over-sampling was conducted by K-SMOTE, and random under-sampling was carried out to reduce the problem scale to form a new dataset. A number of weak classifiers were generated and integration techniques were used to form the final strong classifier. Findings: Experiment was carried out on the UCI imbalanced dataset. The results showed that the proposed algorithm was effective by using the F-value and G-mean value as the evaluation indexes. Originality: In the paper, we improved the SMOTE algorithm and combined over-sampling technology, under-sampling technology and boosting technology to solve the classification problem for imbalanced data. Practical value: The proposed algorithm has important value in imbalanced data classification. It can be applied in the field of different kinds of imbalanced data classification, such as fault detection, intrusion detection, etc. © Liu Ning, 2016.

Yi T.,Shangluo University
Metallurgical and Mining Industry | Year: 2015

In the process of tobacco baking, the temperature and humidity are characteristic of time-varing, large delay and nonlinear, which make the controlling effect of conventional control methods unsatisfying. Recently, the fuzzy control method has been applied in various fields of natural science and social science, and fuzzy control has become an important branch of intelligent control. In this paper, we utilize fuzzy control method, and combine it with the fuzzy math and performance of tobacco-baking, fuzzy language and fuzzy logic rules, designing the fuzzy controller to make the tobacco-control system adaptive and robust. This method realizes the accurate control on the temperature the curing time the tobacco leaves baking, and the system achieve the functions of temperature and humidity automatically check, and the sectional control of temperature and humidity.

Zhao J.,Shangluo University
Open Automation and Control Systems Journal | Year: 2015

The most existing digital watermarking methods were robust to general attacks such as additive noise, image filtering, JPEG compression and so on, but resisting to geometric attacks is still a difficult challenge. This paper proposed a watermarking algorithm resisting to geometrical attacks. The Radon transform was employed to project two dimension image onto projection space. Then the rotation of this image was converted to a translation of the projection. At the same time, the scaling was converted to a scaling of the projection besides the scaling of the projection amplitude. After that, the analytic Fourier–Mellin transform was used to construct invariant moments. These invariant moments are used to design and detect watermark. The experiment demonstrates the proposed method was robust to usually image processing operation, and performed robust to geometrical attacks. © Jie Zhao.

Zhao J.,Shangluo University
Metallurgical and Mining Industry | Year: 2015

In order to improve the robustness and invisibility of image watermarking, this paper proposed a zero watermarking method based on nonsubsampled contourlet transform (NSCT) and Hadamard transform. After doing the nonsubsampled contourlet transform, the low frequent sub band was extracted. This sub band was divided into several blocks. Each block was implemented Hadamard transform. The Hadamard coefficients were used to obtain the construction image. The construction image and watermark were inputs of the cellular neural network, and the zero-watermarking registration image was the output. The image scrambling and cellular neural network improve the security and robustness. To deal with the scaling and rotation attacks, the scale-invariant feature transform (SIFT) was employed. The experimental results demonstrated the proposed method was robust to common image processing. This method is simple, and is convenient for real-time processing.

Kong L.,Shangluo University
Fangzhi Gaoxiao Jichukexue Xuebao | Year: 2015

In real normed linear space, the definition and properties of approximate square R-orthogonality are given. Using the operator theory, it is proved that the approximate square R-orthogonality is approximate B-orthogonality, next the definition of approximate square R-orthogonality preserving mApplng is given. Finally,some sufficient conditions for a bounded linear mApplng to be an approximate square R-orthogonality preserving mApplng are abtained.

Discover hidden collaborations