Fang Gao, Zhangqin Huang, Shulong Wang, Xinrong Ji
Because of the complex architecture and multiple iterations algorithm, neural network is sometimes hard for traditional embedded devices to meet the needs of real-time processing speed in large scale data applications. Manycore processors are directly applicable for parallel implementation of the neural network. In this paper we present a multilayer perception feed forward acceleration framework based on power efficiency manycore processor, including network mapping strategy, data structure design and inter-core communication method. The framework is implemented on a Zynq and Epiphany combined hardware platform with OpenCL. The experimental results show that in a concrete example of character recognition, the framework with Epiphany achieves about four times feed forward acceleration than the dual-core ARM in Zynq with same prediction accuracy level.
{"title":"A Manycore Processor Based Multilayer Perceptron Feedforward Acceleration Framework for Embedded System","authors":"Fang Gao, Zhangqin Huang, Shulong Wang, Xinrong Ji","doi":"10.1109/ICISCE.2016.21","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.21","url":null,"abstract":"Because of the complex architecture and multiple iterations algorithm, neural network is sometimes hard for traditional embedded devices to meet the needs of real-time processing speed in large scale data applications. Manycore processors are directly applicable for parallel implementation of the neural network. In this paper we present a multilayer perception feed forward acceleration framework based on power efficiency manycore processor, including network mapping strategy, data structure design and inter-core communication method. The framework is implemented on a Zynq and Epiphany combined hardware platform with OpenCL. The experimental results show that in a concrete example of character recognition, the framework with Epiphany achieves about four times feed forward acceleration than the dual-core ARM in Zynq with same prediction accuracy level.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89543425","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Liu Yinghui, Wang Qingning, Zhang Donghui, Sun Xiangde, Shen Yang, Xu Xianglian
Nowadays, the backwardness of the power automation management in our country causes the loss of a lot of energy. In order to improve the situation, an anti-stealing mathematical model is introduced in this paper. Firstly, ten factors are selected to build the indictor evaluation system, data mining is used to process lots of the electricity data. Then, a mathematical model based on BP neural network is built for analyzing the customer consumption behavior. With the model, the suspicion coefficient of electricity stealing can be calculated, and the credit rating of power consumer is also classified. In this paper, some typical companies are selected to verify the electricity anti-stealing model, and a conclusion that a feasible idea for the electricity stealing problem is drawn.
{"title":"Research and Application of Electricity Anti-stealing System Based on Neural Network","authors":"Liu Yinghui, Wang Qingning, Zhang Donghui, Sun Xiangde, Shen Yang, Xu Xianglian","doi":"10.1109/ICISCE.2016.224","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.224","url":null,"abstract":"Nowadays, the backwardness of the power automation management in our country causes the loss of a lot of energy. In order to improve the situation, an anti-stealing mathematical model is introduced in this paper. Firstly, ten factors are selected to build the indictor evaluation system, data mining is used to process lots of the electricity data. Then, a mathematical model based on BP neural network is built for analyzing the customer consumption behavior. With the model, the suspicion coefficient of electricity stealing can be calculated, and the credit rating of power consumer is also classified. In this paper, some typical companies are selected to verify the electricity anti-stealing model, and a conclusion that a feasible idea for the electricity stealing problem is drawn.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89724869","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Data Release 12 is the final data release of the SDSS-III, containing all SDSS observations. The massive spectra can not only be used for research of the structure and evolution of the Galaxy but also for multi-waveband identification. In addition, the spectra are a ideal sample for data mining for rare and special objects like white dwarf main-sequence star. WDMS consists of a white dwarf primary and a low-mass main-sequence companion which has positive significance to the study of evolution and parameters of close binaries. In this paper, after feature extraction by PCA, an clustering approach is proposed based on the idea that cluster centers are characterized by a higher density than their neighbors and by a relatively large distance from points with higher densities. A total number of 2,340 WDMS candidates are selected by the method and some of them are new discoveries which prove that our approach of finding special celestial bodies in massive spectra data is feasible.
{"title":"Data Mining for WDMS in SDSS DR12 Archive","authors":"Jiang Bin, Chunyu Ma, W. Wenyu, Wang Wei, Gao Jun","doi":"10.1109/ICISCE.2016.70","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.70","url":null,"abstract":"Data Release 12 is the final data release of the SDSS-III, containing all SDSS observations. The massive spectra can not only be used for research of the structure and evolution of the Galaxy but also for multi-waveband identification. In addition, the spectra are a ideal sample for data mining for rare and special objects like white dwarf main-sequence star. WDMS consists of a white dwarf primary and a low-mass main-sequence companion which has positive significance to the study of evolution and parameters of close binaries. In this paper, after feature extraction by PCA, an clustering approach is proposed based on the idea that cluster centers are characterized by a higher density than their neighbors and by a relatively large distance from points with higher densities. A total number of 2,340 WDMS candidates are selected by the method and some of them are new discoveries which prove that our approach of finding special celestial bodies in massive spectra data is feasible.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87279247","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In [10] and [16], we proposed tools for simultaneous variable selection and parameter estimation in a functional linear model with a functional outcome and a large number of scalar predictor. We call these techniques Function-on-Scalar Lasso (FSL) and Adaptive Function-on-Scalar Lasso(AFSL). A scalar group lasso was used to fit the FSL and AFSL estimates. While this approach works well, we improve it by producing custom ADMM methods which are specifically designed for functional data. We propose this new framework as a computational tool for finding FSL estimates. Through our numerical studies, we demonstrate the computational improvement of our methodology.
{"title":"High-Dimensional Function-on-Scale Regression via the Alternating Direction Method of Multipliers","authors":"Zhaohu Fan, M. Reimherr","doi":"10.1109/ICISCE.2016.93","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.93","url":null,"abstract":"In [10] and [16], we proposed tools for simultaneous variable selection and parameter estimation in a functional linear model with a functional outcome and a large number of scalar predictor. We call these techniques Function-on-Scalar Lasso (FSL) and Adaptive Function-on-Scalar Lasso(AFSL). A scalar group lasso was used to fit the FSL and AFSL estimates. While this approach works well, we improve it by producing custom ADMM methods which are specifically designed for functional data. We propose this new framework as a computational tool for finding FSL estimates. Through our numerical studies, we demonstrate the computational improvement of our methodology.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88285466","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tension and gauge control interact each other intensively in tandem cold rolling process, making control ineffective and unstable. In order to improve the thickness of the strip and tension control accuracy, control system of tandem cold rolling is equipped with the decoupling control of tension and gauge. Applying the method of Neural Network Simple Adaptive Control based on Genetic Algorithm, we design the Neural Network Simple Adaptive Control multi-variable decoupling control system of tension and gauge, The three layer neural network with genetic algorithm is used to decouple the tension and gauge system of cold rolling mill. The simulation result proves that the decoupling control of tension and gauge can be realized, furthermore, the dynamic and static characteristic of control system is better.
{"title":"The Decoupling Control of Tandem Cold Rolling Tension and Gauge","authors":"An Shiqi","doi":"10.1109/ICISCE.2016.248","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.248","url":null,"abstract":"Tension and gauge control interact each other intensively in tandem cold rolling process, making control ineffective and unstable. In order to improve the thickness of the strip and tension control accuracy, control system of tandem cold rolling is equipped with the decoupling control of tension and gauge. Applying the method of Neural Network Simple Adaptive Control based on Genetic Algorithm, we design the Neural Network Simple Adaptive Control multi-variable decoupling control system of tension and gauge, The three layer neural network with genetic algorithm is used to decouple the tension and gauge system of cold rolling mill. The simulation result proves that the decoupling control of tension and gauge can be realized, furthermore, the dynamic and static characteristic of control system is better.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87011845","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Using as few sensor nodes as possible for detecting composite event effectively in large area is a difficult problem, because the composite event contains multiple atomic events which needs many different types of heterogeneous nodes for cooperative monitoring, and the coverage quality would be worse if there are not enough sensor nodes. Most of the traditional methods are focusing on atomic event detection which only needs one type of homogeneous node. Considering the temporal and spatial association, costs and sensing capability of different types of heterogeneous sensor nodes, in this paper, we propose a novel composite event coverage problem with the purpose of minimizing deployment costs subjecting to the constraint of achieving a certain coverage quality, and give a mathematical model for this optimal problem. Then, we propose an exact algorithm and a greedy strategy approximation algorithm to solve this optimization problem. The experimental results and analysis show the performance of the proposed algorithms.
{"title":"Deployment Cost Optimal for Composite Event Detection in Heterogeneous Wireless Sensor Networks","authors":"Xiaoqing Dong","doi":"10.1109/ICISCE.2016.275","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.275","url":null,"abstract":"Using as few sensor nodes as possible for detecting composite event effectively in large area is a difficult problem, because the composite event contains multiple atomic events which needs many different types of heterogeneous nodes for cooperative monitoring, and the coverage quality would be worse if there are not enough sensor nodes. Most of the traditional methods are focusing on atomic event detection which only needs one type of homogeneous node. Considering the temporal and spatial association, costs and sensing capability of different types of heterogeneous sensor nodes, in this paper, we propose a novel composite event coverage problem with the purpose of minimizing deployment costs subjecting to the constraint of achieving a certain coverage quality, and give a mathematical model for this optimal problem. Then, we propose an exact algorithm and a greedy strategy approximation algorithm to solve this optimization problem. The experimental results and analysis show the performance of the proposed algorithms.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87404892","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Among the very high frequency antenna measurement methods used in radio astronomical observation, phase measurement technology is one of vital importance in the process of data acceptance and disposal, for it has great priority in antenna pattern and accuracy in tracking radio source at different region. Particularly, affected by the natural environment, cable phase error is produced, thus the traditional measurement method that depends on different environmental experience value to correct phase error is complex and inaccurate. According to the principle of low frequency antenna array phase measurement, this paper suggests a better method, which uses AD8302 to transfer the two calibration source signal phase difference of the receiver and transfer feeder to the corresponding direct voltage, then realizes single channel AD sampling with 12 bits quantization accuracy, finally does the error analysis and correction to achieve phase error correction. Through testing the precision of phase error of the cable given by the system under measuring conditions of low frequency (2MHz), high frequency (300MHz), etc. We can prove the system has a very high reliability, thus could be widely used in related antenna array error correction and cable error measurement.
{"title":"The Cable Phase Difference Detection of Very High Frequency (VHF) Antenna Array Based on Ad8302","authors":"Linzhu Yue, L. Dong","doi":"10.1109/ICISCE.2016.305","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.305","url":null,"abstract":"Among the very high frequency antenna measurement methods used in radio astronomical observation, phase measurement technology is one of vital importance in the process of data acceptance and disposal, for it has great priority in antenna pattern and accuracy in tracking radio source at different region. Particularly, affected by the natural environment, cable phase error is produced, thus the traditional measurement method that depends on different environmental experience value to correct phase error is complex and inaccurate. According to the principle of low frequency antenna array phase measurement, this paper suggests a better method, which uses AD8302 to transfer the two calibration source signal phase difference of the receiver and transfer feeder to the corresponding direct voltage, then realizes single channel AD sampling with 12 bits quantization accuracy, finally does the error analysis and correction to achieve phase error correction. Through testing the precision of phase error of the cable given by the system under measuring conditions of low frequency (2MHz), high frequency (300MHz), etc. We can prove the system has a very high reliability, thus could be widely used in related antenna array error correction and cable error measurement.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79708621","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Schema matching, aiming at finding the semantic correspondence of concepts between different data sources, has become a hot research area in data integration, data exchange and other areas. Over the past few years, numerous effective schema matching methods have been proposed. In the latest years, researchers have proposed a number of matching methods, which make it possible to identify and discover the semantic correspondence between the data. But schema matching have some challenges in the form of definition, utilization, and combination of element similarity measures, especially in multiple schema matching. In this paper, we study multi-schema matching problem, and convert this issue to indirect matching problem using reference ontology. We find and locate the core word, present composite schema matching methods, and generate matching result. Our experimental results demonstrate the high performance with a well precision and recall.
{"title":"Towards a Composite XML Schema Matching Approach Using Reference Ontology","authors":"Hongjie Fan, Junfei Liu, Kejun Deng","doi":"10.1109/ICISCE.2016.160","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.160","url":null,"abstract":"Schema matching, aiming at finding the semantic correspondence of concepts between different data sources, has become a hot research area in data integration, data exchange and other areas. Over the past few years, numerous effective schema matching methods have been proposed. In the latest years, researchers have proposed a number of matching methods, which make it possible to identify and discover the semantic correspondence between the data. But schema matching have some challenges in the form of definition, utilization, and combination of element similarity measures, especially in multiple schema matching. In this paper, we study multi-schema matching problem, and convert this issue to indirect matching problem using reference ontology. We find and locate the core word, present composite schema matching methods, and generate matching result. Our experimental results demonstrate the high performance with a well precision and recall.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82622777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In order to solve the problem of battlefield targets grouping, based on the research on targets key properties, this paper puts forward a method about using stable center of gravity algorithm based on entropy weight. The K-means clustering algorithms advantages and defects are introduced, and the paper makes improvement to the algorithm for its defects. For the difficulty of Selecting of the initial clustering centers, we put forward the solution of using experience knowledge which means we select the corresponding initial clustering centers according to the specific application scenario. And the research objects of this article are battlefield units. We propose to select the carrier, planes and other important units as the initial clustering centers and establish objects grouping model based on priori knowledge. Later in the experimental section, there is a comparison between the methodwe proposed and the traditional K-means algorithm, and we make a summary which proves the effectiveness of the method proposed.
{"title":"Research on the Algorithm for Grouping Battlefield Targets","authors":"Xiang Ji, J. Hao, Yibin Tu, Hengwei Zhang","doi":"10.1109/ICISCE.2016.124","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.124","url":null,"abstract":"In order to solve the problem of battlefield targets grouping, based on the research on targets key properties, this paper puts forward a method about using stable center of gravity algorithm based on entropy weight. The K-means clustering algorithms advantages and defects are introduced, and the paper makes improvement to the algorithm for its defects. For the difficulty of Selecting of the initial clustering centers, we put forward the solution of using experience knowledge which means we select the corresponding initial clustering centers according to the specific application scenario. And the research objects of this article are battlefield units. We propose to select the carrier, planes and other important units as the initial clustering centers and establish objects grouping model based on priori knowledge. Later in the experimental section, there is a comparison between the methodwe proposed and the traditional K-means algorithm, and we make a summary which proves the effectiveness of the method proposed.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82864218","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the improvement of processor and SDRAM performance, the performance of SDRAM controller becomes the bottleneck of the system performance. In this paper, a Tow-Level Buffered SDRAM controller is proposed, and its design and verification are described. To some extent, the controller improves the throughput of the processor for the SDRAM memory, and provides a solution for the design of high performance system.
{"title":"A Tow-Level Buffered SDRAM Controller","authors":"T. Jin, Wenxin Li, Xiangyu Hu","doi":"10.1109/ICISCE.2016.37","DOIUrl":"https://doi.org/10.1109/ICISCE.2016.37","url":null,"abstract":"With the improvement of processor and SDRAM performance, the performance of SDRAM controller becomes the bottleneck of the system performance. In this paper, a Tow-Level Buffered SDRAM controller is proposed, and its design and verification are described. To some extent, the controller improves the throughput of the processor for the SDRAM memory, and provides a solution for the design of high performance system.","PeriodicalId":6882,"journal":{"name":"2016 3rd International Conference on Information Science and Control Engineering (ICISCE)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2016-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91315752","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：