Guangdong AIB Polytechnic College

Guangzhou, China

Guangdong AIB Polytechnic College

Guangzhou, China
SEARCH FILTERS
Time filter
Source Type

Xin-Yao Z.,Guangdong AIB Polytechnic College | Ruo-He Y.,South China University of Technology | Pei-Xin Q.,Henan Institute of Science and Technology
International Journal of Advancements in Computing Technology | Year: 2012

This paper presents a life prediction method based on least squares support vector machine(LSSVM) for small data samples from Weibull distributions. This method can directly predict lifetime by building a least squares support vector regression model. Monte Carlo simulations demonstrated the effectiveness of LSSVM approach in prediction of Weibull data sets. Error back propagation (BP) neural network was also compared with it. The simulation results show that LSSVM method has better generalization and higher accuracy of life prediction than BP neural network when dealing with small data samples from Weibull distribution.


Xin-Yao Z.,Guangdong AIB Polytechnic College
Lecture Notes in Electrical Engineering | Year: 2014

To resolve reliability evaluation for integrated circuit with small sample, a lifetime predictionmethod is presented based on least squares support vector machine (LSSVM) for small data samples when the failure distributions are assumed to be exponential distributions. Based on the known failure data, a LSSVM regression model can be constructed, and the unknown lifetime can be evaluated by the LSSVM model.MonteCarlo simulations demonstrated the effectiveness of LSSVMapproach in prediction of exponential datasets. Error back propagation (BP) neural network was also compared with it. The simulation results show that LSSVM method has better generalization and higher accuracy of life prediction than BP neural network when dealing with small data samples from exponential distribution. © Springer-Verlag Berlin Heidelberg 2014.


Zou X.-Y.,Guangdong AIB Polytechnic College
Lecture Notes in Electrical Engineering | Year: 2014

Normal distribution is often used to describe the failure characteristic of integrated circuit (IC). It is difficult to obtain a lot of reliability data samples for microelectronics devices in nano era. To resolve this problem, a lifetime prediction method based on least squares support vector machine (LSSVM) is presented for small data samples when the failure distributions are assumed to be normal distributions. Monte Carlo simulations demonstrated the effectiveness of LSSVM approach in prediction of normal datasets. Error back propagation (BP) neural network was also compared with it. The simulation results show that linear kernel is better than RBF kernel for LSSVM approach, and LSSVM method has better generalization and higher accuracy of life prediction than BP neural network when dealing with small data samples from normal distribution. © Springer-Verlag Berlin Heidelberg 2014.


Li F.,Guangdong AIB Polytechnic College
Advances in Intelligent Systems and Computing | Year: 2013

Constructing the network courses based on personalized learning is an important way to satisfy the learners' individual demands and improve network teaching qualities. Taking Applied Technology of Single Chip Processor as the example, this paper analyzes the present problems of network courses, describes the features of network courses on the basis of personalized learning and discusses the basic frames of network courses purposely realizing personal learning. In addition, it provides related algorithms using to design network courses on the foundation of personalized learning, regarding the Web mine technology and ontology technology as the crucial ones. © 2013 Springer-Verlag.


Li F.,Guangdong AIB Polytechnic College
Advances in Intelligent Systems and Computing | Year: 2013

Along with the rapid development and the widespread application of the computer technology and Internet technology, education and teaching have gradually entered the era of computer networks. The intelligent network teaching system has incomparable superiority of the traditional teaching system, which not only standardize the management of education and teaching, but also reduce the workload of the education and teaching. What's more, it can greatly improve the efficiency of teaching management. Research of the Intelligent Network Teaching System from the current and future development trend analysis and summary will explore the specific implementation steps and the application of this new type of intelligent network teaching system. © 2013 Springer-Verlag.


Zou X.-Y.,Guangdong AIB Polytechnic College
Lecture Notes in Electrical Engineering | Year: 2012

Markov model has been used to model user navigational behavior on the World Wide Web. Most existing Markov models based on recommendation algorithms make recommendation just through Web server logs without Web topology. In this chapter, a time-space hybrid Markov model (TSHMM) recommendation algorithm is proposed. First we build both time model from Web log and space model from Web topology respectively. Then we build time-space hybrid model by combining time model with space model. Hybrid model can make recommendation according to user preferences. Some experiments are conducted to validate the hybrid model. The experimental results show that hybrid model can be successfully applied to Web recommendation. © 2012 Springer Science+Business Media B.V.


Zheng J.,Guangdong AIB Polytechnic College
International Journal of Security and its Applications | Year: 2016

Teaching quality evaluation usually involves multiple influence factors with grey and uncertain information. This brings about limitations to effective evaluation of teaching quality in colleges and universities. Thus, this paper draws merits from previous research and proposes a teaching quality evaluation system and model for colleges and universities. First, based on real situations, grey categories of evaluation state for teaching quality are established. The definite weighted functions of grey category of evaluation state are confirmed. Specific steps of the teaching quality evaluation model based on grey clustering analysis are accounted for. Finally, a case study is introduced to verify the model. This model enlightens a new way to evaluate teaching quality in colleges and universities. © 2016 SERSC.


Li X.,Guangdong AIB Polytechnic College
Proceedings - 5th International Conference on Intelligent Networking and Collaborative Systems, INCoS 2013 | Year: 2013

While the traditional single-processor system can not meet the requirements of modern array signal system, a new high-speed parallel system based on DSPs and FPGA is designed. The system complies with VITA 46 standard. The paper introduces the system design in detail, including system architecture, DSP subsystem and FPGA subsystem. The system based on TMS320C6678 can apply widely in the voice processing, sonar, radar, communication, real-time image processing and other high-speed signal processing fields. © 2013 IEEE.


Li X.,Guangdong AIB Polytechnic College
Proceedings - 2013 8th International Conference on P2P, Parallel, Grid, Cloud and Internet Computing, 3PGCIC 2013 | Year: 2013

As the existing block-wise lossless image authentication algorithms sacrifice the precision of locating distortions, a new lossless image authentication algorithm based on three-compa image basic blocks is proposed. The algorithm trisects the original image recursively in order to get basic blocks. When the watermark can't be embedded into the authentication block, the algorithm tries to merge only one basic block into current authentication block. As a result, each authentication block's size is expanded at a minimum level. Moreover, in order to prevent vector quantization attack, a digital signature algorithm is taken and signature is repeatedly embedded into the authentication block via a lossless watermarking algorithm. Theoretical analysis and experimental results demonstrate that the algorithm is more accurate than the algorithm based on adaptive combinations of image basic blocks. © 2013 IEEE.


Xiangyang L.,Guangdong AIB Polytechnic College
Proceedings - 4th International Conference on Emerging Intelligent Data and Web Technologies, EIDWT 2013 | Year: 2013

Low-Voltage Differential Signaling (LVDS) evolves over the last decade to meet specific requirements such as Bus LVDS and Multipoint LVDS. This paper first presents a method to implement the multi-channel LVDS in the parallel RapidIO protocol using CPLD device and VHDL language, followed by a detailed discussion of the data line transmission error generated during the process of high-speed data transportation due to clock-data skew and difference between transmission lines. A logical component, 4-bit-channel aligner, is developed and simulated to solve this sort of transmission error. Finally, an evaluation board of RapidIO protocol is developed to evaluate the correctness of multi-channel LVDS data transmission. Some experiments show that the multi-channel LVDS is simple and strong anti-interference. © 2013 IEEE.

Loading Guangdong AIB Polytechnic College collaborators
Loading Guangdong AIB Polytechnic College collaborators