Extreme learning machine (ELM) is a new type of feed forward neural network. Compared with traditional single hidden layer feed forward neural networks, ELM executes with higher training speed and produces smaller error. Due to random input weights and hidden biases, ELM might need numerous hidden neurons to achieve a reasonable accuracy. A new ELM learning algorithm, which was optimized by the Firefly Algorithm (FA), was proposed in this paper. FA was used to select the input weights and biases of hidden layer, and then the output weights could be calculated. To test the validity of proposed method, a simulation experiments about the approximation curves of the SINC function was done. The results showed that the proposed algorithm achieved better performance with less hidden neurons than other similar methods.
{"title":"A New Extreme Learning Machine Optimized by Firefly Algorithm","authors":"Qiang Zhang, Hongxin Li, Changnian Liu, Wei Hu","doi":"10.1109/ISCID.2013.147","DOIUrl":"https://doi.org/10.1109/ISCID.2013.147","url":null,"abstract":"Extreme learning machine (ELM) is a new type of feed forward neural network. Compared with traditional single hidden layer feed forward neural networks, ELM executes with higher training speed and produces smaller error. Due to random input weights and hidden biases, ELM might need numerous hidden neurons to achieve a reasonable accuracy. A new ELM learning algorithm, which was optimized by the Firefly Algorithm (FA), was proposed in this paper. FA was used to select the input weights and biases of hidden layer, and then the output weights could be calculated. To test the validity of proposed method, a simulation experiments about the approximation curves of the SINC function was done. The results showed that the proposed algorithm achieved better performance with less hidden neurons than other similar methods.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132837742","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 the case of metrics-based software defect prediction, an intelligent selection of metrics plays an important role in improving the model performance. In this paper, we use different ways for feature selection and dimensionality reduction to determine the most important software metrics. Three different classifiers are utilized, namely Naïve Bayes, support vector machine and decision tree. On the publicly NASA data, a comparative experiment results show that instead of 22 or more metrics, less than 10 metrics can get better performance.
{"title":"A Study on the Significance of Software Metrics in Defect Prediction","authors":"Ye Xia, G. Yan, Qianran Si","doi":"10.1109/ISCID.2013.199","DOIUrl":"https://doi.org/10.1109/ISCID.2013.199","url":null,"abstract":"In the case of metrics-based software defect prediction, an intelligent selection of metrics plays an important role in improving the model performance. In this paper, we use different ways for feature selection and dimensionality reduction to determine the most important software metrics. Three different classifiers are utilized, namely Naïve Bayes, support vector machine and decision tree. On the publicly NASA data, a comparative experiment results show that instead of 22 or more metrics, less than 10 metrics can get better performance.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133165178","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}
QoS multicast routing problem is a nonlinear combination optimization problem, which is difficult to get the global solution by using the traditional algorithm. In this paper, we use the Firefly Algorithm (FA) to solve the multi-constrained QoS multicast routing problem and QoS-FA algorithm is presented. FA is a novel heuristic stochastic algorithm and has been applied to many optimization fields. The simulation results show that the QoS-FA algorithm can search the optical multicast tree and has better performance.
{"title":"QoS Multicast Routing Based on Firefly Algorithm","authors":"Jie Yuan, Yafei Tian, Shan Wang, Changnian Liu","doi":"10.1109/ISCID.2013.47","DOIUrl":"https://doi.org/10.1109/ISCID.2013.47","url":null,"abstract":"QoS multicast routing problem is a nonlinear combination optimization problem, which is difficult to get the global solution by using the traditional algorithm. In this paper, we use the Firefly Algorithm (FA) to solve the multi-constrained QoS multicast routing problem and QoS-FA algorithm is presented. FA is a novel heuristic stochastic algorithm and has been applied to many optimization fields. The simulation results show that the QoS-FA algorithm can search the optical multicast tree and has better performance.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"88 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132630661","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}
Cloud has been rapidly and broadly discussed in the last several years, definitely it's been carried into practice by many enterprises due to its undoubted expectation. Plenty of corporations (IT or Non-IT) have chosen to provide their services through cloud. To reconsider providing the same or better services through cloud, the corporations will have to implement the reengineering of all their software services. During reengineering, how to rapidly migrate the applications into cloud and how to reconstruct the development framework and architecture is the key point of success, which have not be deeply researched over by the cloud vendors yet. So this paper proposes a rapid cloud migration solution (RCMS) which includes advanced migration strategy, complete security framework, rapid development process, highly automatic reengineering, guaranteed performance monitoring and flexible storage services. These features make the migration rapid and stable which are proved cost saving and easy-maintenance.
{"title":"RCMS: Rapid Cloud Migration Solution","authors":"Hexin Lv, Jingjing Liu","doi":"10.1109/ISCID.2013.112","DOIUrl":"https://doi.org/10.1109/ISCID.2013.112","url":null,"abstract":"Cloud has been rapidly and broadly discussed in the last several years, definitely it's been carried into practice by many enterprises due to its undoubted expectation. Plenty of corporations (IT or Non-IT) have chosen to provide their services through cloud. To reconsider providing the same or better services through cloud, the corporations will have to implement the reengineering of all their software services. During reengineering, how to rapidly migrate the applications into cloud and how to reconstruct the development framework and architecture is the key point of success, which have not be deeply researched over by the cloud vendors yet. So this paper proposes a rapid cloud migration solution (RCMS) which includes advanced migration strategy, complete security framework, rapid development process, highly automatic reengineering, guaranteed performance monitoring and flexible storage services. These features make the migration rapid and stable which are proved cost saving and easy-maintenance.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115172011","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}
We address an integrated logistics system where decisions on location of depot, vehicle routing are considered simultaneously. Total cost and service quality are common criteria influencing decision-making. Literature on location routing problem (LRP) addressed the location and vehicle routing decisions with a common assumption that each vehicle can only performance pickup or delivery assignment in each dispatch. However, both demands of each customer often require be satisfied at the same time. In this paper we consider a LRP with simultaneous pickup and delivery to minimize total cost and customer waiting time. We formulate a nonlinear multi-objective integrated programming model for the problem. A heuristic algorithm based on tabu search is proposed to solve the large-size problem. We then empirically evaluate the strengths of the proposed formulations with respect to their ability to find optimal solutions or strong lower bounds, and investigate the effectiveness of the proposed heuristic approach. Results show that the proposed heuristic approach is computationally efficient in finding good quality solutions.
{"title":"Multi-objective Metaheuristics for a Location-Routing Problem with Simultaneous Pickup and Delivery","authors":"Xuefeng Wang","doi":"10.1109/ISCID.2013.197","DOIUrl":"https://doi.org/10.1109/ISCID.2013.197","url":null,"abstract":"We address an integrated logistics system where decisions on location of depot, vehicle routing are considered simultaneously. Total cost and service quality are common criteria influencing decision-making. Literature on location routing problem (LRP) addressed the location and vehicle routing decisions with a common assumption that each vehicle can only performance pickup or delivery assignment in each dispatch. However, both demands of each customer often require be satisfied at the same time. In this paper we consider a LRP with simultaneous pickup and delivery to minimize total cost and customer waiting time. We formulate a nonlinear multi-objective integrated programming model for the problem. A heuristic algorithm based on tabu search is proposed to solve the large-size problem. We then empirically evaluate the strengths of the proposed formulations with respect to their ability to find optimal solutions or strong lower bounds, and investigate the effectiveness of the proposed heuristic approach. Results show that the proposed heuristic approach is computationally efficient in finding good quality solutions.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115243724","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}
The network traffic prediction model is the foundation of network performance analysis and designing. Aiming at limitation of the conventional network traffic time series prediction model and the problem that BP algorithms easily plunge into local solution, an optimization algorithm-PSO-QI which combine particle swarm optimization (PSO) and the quantum principle is proposed, and can alleviate the premature convergence validly. Then, the parameters of BP neural network were optimized and the time series of network traffic data was modeled and forecasted based on BP neural network and PSO-QI. Experiments showed that PSOQI-BP neural network has better precision and adaptability compared with the traditional neural network.
{"title":"A Network Traffic Prediction Model Based on Quantum Inspired PSO and Neural Network","authors":"Kun Zhang, L. Liang, Ying Huang","doi":"10.1109/ISCID.2013.168","DOIUrl":"https://doi.org/10.1109/ISCID.2013.168","url":null,"abstract":"The network traffic prediction model is the foundation of network performance analysis and designing. Aiming at limitation of the conventional network traffic time series prediction model and the problem that BP algorithms easily plunge into local solution, an optimization algorithm-PSO-QI which combine particle swarm optimization (PSO) and the quantum principle is proposed, and can alleviate the premature convergence validly. Then, the parameters of BP neural network were optimized and the time series of network traffic data was modeled and forecasted based on BP neural network and PSO-QI. Experiments showed that PSOQI-BP neural network has better precision and adaptability compared with the traditional neural network.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123872846","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}
This paper proposes a novel classification method named KF-SVM (Kernel Fisher, Support Vector Machine), which is used for the EEG (Electroencephalography) classification of two classes of imagery data in BCIs (brain-computer interfaces). This method combines the kernel fisher and SVM. Its detailed process is as follows: First, the CSP (Common Spatial Patterns) is used to obtain features, and then the within-class scatter is calculated based on these features. The scatter is added into the RBF (Radical Basis Function) kernel function to construct a new kernel function. The obtained new kernel is integrated into the support vector machine to get a new classification model. The KF-SVM may overcome the following defects of the SVM: 1) the SVM maximizes the classification margin without considering within-class scatter. 2) The classification surface of the SVM between two types of EEG data only depends on boundary samples and misclassified samples. To evaluate effectiveness of the proposed KF-SVM method, the data from the 2008 international BCI competition and experiments of our laboratory are processed. The experimental result shows that the proposed KF-SVM classification algorithm can well classify EEG data and improve the correct rate of EEG recognition in BCIs.
{"title":"A New Classification Method Based on KF-SVM in Brain Computer Interfaces","authors":"Yang Banghua, Han Zhijun, Wang Qian, He Liangfei","doi":"10.1109/ISCID.2013.55","DOIUrl":"https://doi.org/10.1109/ISCID.2013.55","url":null,"abstract":"This paper proposes a novel classification method named KF-SVM (Kernel Fisher, Support Vector Machine), which is used for the EEG (Electroencephalography) classification of two classes of imagery data in BCIs (brain-computer interfaces). This method combines the kernel fisher and SVM. Its detailed process is as follows: First, the CSP (Common Spatial Patterns) is used to obtain features, and then the within-class scatter is calculated based on these features. The scatter is added into the RBF (Radical Basis Function) kernel function to construct a new kernel function. The obtained new kernel is integrated into the support vector machine to get a new classification model. The KF-SVM may overcome the following defects of the SVM: 1) the SVM maximizes the classification margin without considering within-class scatter. 2) The classification surface of the SVM between two types of EEG data only depends on boundary samples and misclassified samples. To evaluate effectiveness of the proposed KF-SVM method, the data from the 2008 international BCI competition and experiments of our laboratory are processed. The experimental result shows that the proposed KF-SVM classification algorithm can well classify EEG data and improve the correct rate of EEG recognition in BCIs.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125448961","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}
Load balancing technology can solve the network congestion problems of modern network which is caused by uneven distribution of traffic. As the network link load balancing is an NP-complete problem, it is difficult to use traditional method to deal with, introducing the idea of genetic algorithm. Using genetic algorithm, the characteristics of efficient and parallel can help to find the global optimal solution quickly. Article on the basis of traditional genetic algorithm, this paper puts forward a network link load balancing strategy based on improved genetic algorithm. Experiments show that it can find the answer to the problem better.
{"title":"A Study of Link Load Balancing Based on Improved Genetic Algorithm","authors":"Li Zhao, Yu-min Dong, Chen-yang Huang","doi":"10.1109/ISCID.2013.183","DOIUrl":"https://doi.org/10.1109/ISCID.2013.183","url":null,"abstract":"Load balancing technology can solve the network congestion problems of modern network which is caused by uneven distribution of traffic. As the network link load balancing is an NP-complete problem, it is difficult to use traditional method to deal with, introducing the idea of genetic algorithm. Using genetic algorithm, the characteristics of efficient and parallel can help to find the global optimal solution quickly. Article on the basis of traditional genetic algorithm, this paper puts forward a network link load balancing strategy based on improved genetic algorithm. Experiments show that it can find the answer to the problem better.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"291 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122717966","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}
T. Peng, Delong Zhang, Xiaoming Liu, Shang Wang, Wanli Zuo
Most researches on co-authorship network analyze the author's information globally according to the overall network topology structure, instead of analyzing the author's local network. Therefore, this paper presents a community mining algorithm and divides big co-authorship network into small communities, in which entities' relationship is closer. Then we mine central authors in community by three different centrality standards including closeness centrality, eigenvector centrality and a new proposed measure termed extensity degree centrality. We choose the SIGMOD data as datasets and measure the centrality from different views. And experiments in co-authorship network achieve many interesting results, which indicate our technique is efficient and feasible, and also have reference value for scientific evaluation.
{"title":"Central Author Mining from Co-authorship Network","authors":"T. Peng, Delong Zhang, Xiaoming Liu, Shang Wang, Wanli Zuo","doi":"10.1109/ISCID.2013.64","DOIUrl":"https://doi.org/10.1109/ISCID.2013.64","url":null,"abstract":"Most researches on co-authorship network analyze the author's information globally according to the overall network topology structure, instead of analyzing the author's local network. Therefore, this paper presents a community mining algorithm and divides big co-authorship network into small communities, in which entities' relationship is closer. Then we mine central authors in community by three different centrality standards including closeness centrality, eigenvector centrality and a new proposed measure termed extensity degree centrality. We choose the SIGMOD data as datasets and measure the centrality from different views. And experiments in co-authorship network achieve many interesting results, which indicate our technique is efficient and feasible, and also have reference value for scientific evaluation.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128839197","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}
Zheng Shou, Binqiang Yu, Gang Chen, Hengjin Cai, Qiaochu Liu
Online 3D Virtual Garment Try-on System is deeply needed and would be quite popular if we could improve its accuracy, effect, and user experience. In order to achieve these goals, we propose several key designs in implementing it. In this paper, we discuss the system architecture design at first. Then we introduce a simple but effective method to model 3D body prototypes. Based on models in Poser software, we segment them into layers and then do triangularizition to get triangular surfaces. As for garment modeling based on Spring-Mass model and physical try-on simulation, we propose a novel method based on uniform grid to detect collision. Store references of triangular surfaces into grids that they occupied and then calculate which grids a moving line of mass go through. Get triangular surfaces in these grids out and then judge whether the moving line intersects with them. It achieved fast detection in around 0.1% of time consuming by using linear searching. The performances of body modeling and try-on simulation are satisfying, and real-time responses could be achieved because of less complex computation and light-scale data transformation.
{"title":"Key Designs in Implementing Online 3D Virtual Garment Try-On System","authors":"Zheng Shou, Binqiang Yu, Gang Chen, Hengjin Cai, Qiaochu Liu","doi":"10.1109/ISCID.2013.46","DOIUrl":"https://doi.org/10.1109/ISCID.2013.46","url":null,"abstract":"Online 3D Virtual Garment Try-on System is deeply needed and would be quite popular if we could improve its accuracy, effect, and user experience. In order to achieve these goals, we propose several key designs in implementing it. In this paper, we discuss the system architecture design at first. Then we introduce a simple but effective method to model 3D body prototypes. Based on models in Poser software, we segment them into layers and then do triangularizition to get triangular surfaces. As for garment modeling based on Spring-Mass model and physical try-on simulation, we propose a novel method based on uniform grid to detect collision. Store references of triangular surfaces into grids that they occupied and then calculate which grids a moving line of mass go through. Get triangular surfaces in these grids out and then judge whether the moving line intersects with them. It achieved fast detection in around 0.1% of time consuming by using linear searching. The performances of body modeling and try-on simulation are satisfying, and real-time responses could be achieved because of less complex computation and light-scale data transformation.","PeriodicalId":297027,"journal":{"name":"2013 Sixth International Symposium on Computational Intelligence and Design","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2013-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131093385","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：