Classifier ensembles have been shown to outperform single classifier systems. An apparent necessary condition for ensembles to outperform single systems is that the classifier systems exhibit a reasonable degree of "diversity". It has also been demonstrated that diversity is an important predictive factor for the improvement. However, in lack of a universally accepted definition, various diversity measures have been proposed and applied in the literature. A natural question then follows: How can we compare, and hence choose among, various diversity measures? This work exploits analytically the relationships among several well-accepted diversity measures. These different diversity measures are proved to be closely related, which facilitates further research on classifier ensembles since the effective number of diversity measures is reduced by such close relationships.
{"title":"On the Relationships Among Various Diversity Measures in Multiple Classifier Systems","authors":"Y. Chung, D. Hsu, C. Tang","doi":"10.1109/I-SPAN.2008.46","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.46","url":null,"abstract":"Classifier ensembles have been shown to outperform single classifier systems. An apparent necessary condition for ensembles to outperform single systems is that the classifier systems exhibit a reasonable degree of \"diversity\". It has also been demonstrated that diversity is an important predictive factor for the improvement. However, in lack of a universally accepted definition, various diversity measures have been proposed and applied in the literature. A natural question then follows: How can we compare, and hence choose among, various diversity measures? This work exploits analytically the relationships among several well-accepted diversity measures. These different diversity measures are proved to be closely related, which facilitates further research on classifier ensembles since the effective number of diversity measures is reduced by such close relationships.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125011239","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}
Parallel computing is notoriously challenging due to the difficulty in developing correct and efficient programs. With the arrival of multi-core processors for desktop systems, desktop applications must now be parallelised. However achieving task parallelism for such object-oriented programs has traditionally been, and still remains, difficult. This paper presents a powerful task concept for parallel object-oriented programming and presents the results from a source-to-source compiler and runtime system. With the addition of a single keyword, the sequential code does not require restructuring and asynchronous task management is performed on behalf of the programmer; the parallel code required to realise task parallelism looks very much like the sequential counterpart. An intuitive solution is provided to handle task dependencies as well as integrating different task concepts into one model.
{"title":"Task Parallelism for Object Oriented Programs","authors":"Nasser Giacaman, O. Sinnen","doi":"10.1109/I-SPAN.2008.32","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.32","url":null,"abstract":"Parallel computing is notoriously challenging due to the difficulty in developing correct and efficient programs. With the arrival of multi-core processors for desktop systems, desktop applications must now be parallelised. However achieving task parallelism for such object-oriented programs has traditionally been, and still remains, difficult. This paper presents a powerful task concept for parallel object-oriented programming and presents the results from a source-to-source compiler and runtime system. With the addition of a single keyword, the sequential code does not require restructuring and asynchronous task management is performed on behalf of the programmer; the parallel code required to realise task parallelism looks very much like the sequential counterpart. An intuitive solution is provided to handle task dependencies as well as integrating different task concepts into one model.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123742582","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 this paper, we propose an efficient algorithm that finds disjoint paths for node-to-set routing in dual-cube. Dual-cube is a hypercube-like interconnection network with about half of links per node compared with the hypercube containing equal number of nodes. For a dual-cube Dn with n links per node, the algorithm finds n disjoint paths, s rarr ti, 1 les i les n, in 0(n2 log n) time and the maximum length of the paths is bounded by 3n + 3.
{"title":"Node-to-Set Disjoint Paths Routing in Dual-Cube","authors":"K. Kaneko, S. Peng","doi":"10.1109/I-SPAN.2008.18","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.18","url":null,"abstract":"In this paper, we propose an efficient algorithm that finds disjoint paths for node-to-set routing in dual-cube. Dual-cube is a hypercube-like interconnection network with about half of links per node compared with the hypercube containing equal number of nodes. For a dual-cube Dn with n links per node, the algorithm finds n disjoint paths, s rarr ti, 1 les i les n, in 0(n2 log n) time and the maximum length of the paths is bounded by 3n + 3.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127648179","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}
Chung-Ming Huang, Chao-Hsien Lee, Jun-Hao Zhuang, Tsung-Yi Tsai, Chi-Chang Liu
Proxy servers are widespread to be adopted in network applications. Applying proxies over wireless networks not only uses the cached data efficiently but also improves transmission quality effectively. However, if a mobile node (MN) keeps changing its location and getting streaming service simultaneously, it is unreasonable that the MN connects to the same proxy for receiving data all the way. In order to resolve this problem, an application-layer proxy handoff is defined to help MNs switch their proxy servers dynamically. In the current network environment, users can access the Internet via different Internet Service Providers (ISPs). Regarding the traversal among the network domains belonging to different ISPs, this paper proposes a SIP-based proxy handoff mechanism to help the cross-domain proxy switching process. Besides, the proposed mechanism takes multimedia streaming services into consideration and employs MPEG-4 layering technique to smooth the transition from one proxy to another one.
{"title":"SIP-Based Cross-Domain Proxy Handoff for Mobile Streaming Services","authors":"Chung-Ming Huang, Chao-Hsien Lee, Jun-Hao Zhuang, Tsung-Yi Tsai, Chi-Chang Liu","doi":"10.1109/I-SPAN.2008.21","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.21","url":null,"abstract":"Proxy servers are widespread to be adopted in network applications. Applying proxies over wireless networks not only uses the cached data efficiently but also improves transmission quality effectively. However, if a mobile node (MN) keeps changing its location and getting streaming service simultaneously, it is unreasonable that the MN connects to the same proxy for receiving data all the way. In order to resolve this problem, an application-layer proxy handoff is defined to help MNs switch their proxy servers dynamically. In the current network environment, users can access the Internet via different Internet Service Providers (ISPs). Regarding the traversal among the network domains belonging to different ISPs, this paper proposes a SIP-based proxy handoff mechanism to help the cross-domain proxy switching process. Besides, the proposed mechanism takes multimedia streaming services into consideration and employs MPEG-4 layering technique to smooth the transition from one proxy to another one.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"107 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127330175","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}
A network intrusion detection system (NIDS) monitors all network actions and generates alarms when it detects suspicious attempts. We present a data mining technique to assist network administrators to analyze and reduce false positive alarms that are produced by a NIDS. Our data mining technique is based on a growing hierarchical self-organizing map (GHSOM) that adjusts its architecture during an unsupervised training process according to the characteristics of the input alarm data. GHSOM clusters these alarms in a way that supports network administrators in making decisions about true and false alarms. Our empirical results show that our technique is useful for real-world intrusion data.
{"title":"Growing Hierarchical Self-Organizing Map for Filtering Intrusion Detection Alarms","authors":"Maya Shehab, N. Mansour, Ahmad Faour","doi":"10.1109/I-SPAN.2008.42","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.42","url":null,"abstract":"A network intrusion detection system (NIDS) monitors all network actions and generates alarms when it detects suspicious attempts. We present a data mining technique to assist network administrators to analyze and reduce false positive alarms that are produced by a NIDS. Our data mining technique is based on a growing hierarchical self-organizing map (GHSOM) that adjusts its architecture during an unsupervised training process according to the characteristics of the input alarm data. GHSOM clusters these alarms in a way that supports network administrators in making decisions about true and false alarms. Our empirical results show that our technique is useful for real-world intrusion data.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131676787","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}
Recently, routing with disjoint paths has received much attention to provide more performance and fault tolerance. On the other hand, WK-recursive mesh network has been studied extensively due to its favorable properties such as high degree of stability and resilience. The connectivity of a network is an important measure of fault tolerance, while the diameter represents the worst-case transmission delay between two arbitrary nodes. In this paper, we propose one-to-one and one-to-many disjoint routings mechanisms for WK-recursive mesh networks to further exhibit the performance and fault tolerance of such networks.
{"title":"One-to-one and One-to-many node-disjoint Routing Algorithms for WK-Recursive networks","authors":"Saeedeh Bakhshi, H. Sarbazi-Azad","doi":"10.1109/I-SPAN.2008.37","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.37","url":null,"abstract":"Recently, routing with disjoint paths has received much attention to provide more performance and fault tolerance. On the other hand, WK-recursive mesh network has been studied extensively due to its favorable properties such as high degree of stability and resilience. The connectivity of a network is an important measure of fault tolerance, while the diameter represents the worst-case transmission delay between two arbitrary nodes. In this paper, we propose one-to-one and one-to-many disjoint routings mechanisms for WK-recursive mesh networks to further exhibit the performance and fault tolerance of such networks.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"155 10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128665488","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}
Cheng-Kuan Lin, Jimmy J. M. Tan, Hua-Min Huang, D. Hsu, Lih-Hsing Hsu
A hamiltonian cycle C of a graph G is described as langu1, u2,..., un(G), u1rang to emphasize the order of vertices in C. Thus, u1 is the start vertex and ui is the i-th vertex in C. Two hamiltonian cycles of G start at a vertex x, C1 = langu1, u2,..., un(G), u1rang and C2 = langv1, v2,..., vn(G), v1rang, are independent if x = u1 = v1 and u1 ne vi for every i, 2 les i les n(G). A set of hamiltonian cycles {C1, C2,..., Ck} of G are mutually independent if any two different hamiltonian cycles are independent. The mutually independent hamiltonicity of graph G, IHC(G), is the maximum integer k such that for any vertex u of G there exist k-mutually independent hamiltonian cycles ofG starting at u. Inthispaper, we are going to study IHC(G) for the n-dimensional pancake graph Pn and the n-dimensional star graph Sn. We prove that IHC(Pn) = n - 1 if n ges 4 and IHC(Sn) = n-1 if nges5.
{"title":"Mutually Independent Hamiltonianicity of Pancake Graphs and Star Graphs","authors":"Cheng-Kuan Lin, Jimmy J. M. Tan, Hua-Min Huang, D. Hsu, Lih-Hsing Hsu","doi":"10.1109/I-SPAN.2008.41","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.41","url":null,"abstract":"A hamiltonian cycle C of a graph G is described as langu<sub>1</sub>, u<sub>2</sub>,..., u<sub>n(G)</sub>, u<sub>1</sub>rang to emphasize the order of vertices in C. Thus, u<sub>1</sub> is the start vertex and u<sub>i</sub> is the i-th vertex in C. Two hamiltonian cycles of G start at a vertex x, C<sub>1</sub> = langu<sub>1</sub>, u<sub>2</sub>,..., u<sub>n(G)</sub>, u<sub>1</sub>rang and C<sub>2</sub> = langv<sub>1</sub>, v<sub>2</sub>,..., v<sub>n(G)</sub>, v<sub>1</sub>rang, are independent if x = u<sub>1</sub> = v<sub>1</sub> and u<sub>1</sub> ne v<sub>i</sub> for every i, 2 les i les n(G). A set of hamiltonian cycles {C<sub>1</sub>, C<sub>2</sub>,..., C<sub>k</sub>} of G are mutually independent if any two different hamiltonian cycles are independent. The mutually independent hamiltonicity of graph G, IHC(G), is the maximum integer k such that for any vertex u of G there exist k-mutually independent hamiltonian cycles ofG starting at u. Inthispaper, we are going to study IHC(G) for the n-dimensional pancake graph P<sub>n</sub> and the n-dimensional star graph S<sub>n</sub>. We prove that IHC(P<sub>n</sub>) = n - 1 if n ges 4 and IHC(S<sub>n</sub>) = n-1 if nges5.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"121 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128498057","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. Kueng, Tyne Liang, Jimmy J. M. Tan, Lih-Hsing Hsu
Faults in a network may take various forms such as hardware/software errors, node/link faults, etc. In this paper, node-faults are addressed. Let F be a faulty set of f les 2n - 6 conditional node-faults in an injured n-cube Qn such that every node of Qn still has at least two fault - free neighbors. Then we show that Qn - F contains a path of length at least 2n - 2f - 1 (respectively, 2n - 2f - 2) between any two nodes of odd (respectively, even) distance. Since an n-cube is a bipartite graph, such kind of the fault- free path turns out to be the longest one in the case when all faulty nodes belong to the same partite set.
{"title":"On the Longest Fault-Free Paths in Hypercubes with More Faulty Nodes","authors":"T. Kueng, Tyne Liang, Jimmy J. M. Tan, Lih-Hsing Hsu","doi":"10.1109/I-SPAN.2008.29","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.29","url":null,"abstract":"Faults in a network may take various forms such as hardware/software errors, node/link faults, etc. In this paper, node-faults are addressed. Let F be a faulty set of f les 2n - 6 conditional node-faults in an injured n-cube Qn such that every node of Qn still has at least two fault - free neighbors. Then we show that Qn - F contains a path of length at least 2n - 2f - 1 (respectively, 2n - 2f - 2) between any two nodes of odd (respectively, even) distance. Since an n-cube is a bipartite graph, such kind of the fault- free path turns out to be the longest one in the case when all faulty nodes belong to the same partite set.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132404800","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}
Several bandwidth management systems have been developed to reserve, in advance, dedicated connections for high-performance applications. We describe the in-advance reservation capabilities of these systems as well as the bandwidth scheduling and path computation algorithms used. An analytical and experimental evaluation of these algorithms also is provided. Our experiments indicate that for the fixed-slot problem, the minimum-hop feasible path algorithm proposed by us in [8] maximizes network utilization for large networks while the dynamic adaptive feasible path algorithm proposed in this paper does this for small networks.
{"title":"An Evaluation of In-Advance Bandwidth Scheduling Algorithms for Connection-Oriented Networks","authors":"Eun-Sung Jung, Yan Li, S. Ranka, S. Sahni","doi":"10.1109/I-SPAN.2008.34","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.34","url":null,"abstract":"Several bandwidth management systems have been developed to reserve, in advance, dedicated connections for high-performance applications. We describe the in-advance reservation capabilities of these systems as well as the bandwidth scheduling and path computation algorithms used. An analytical and experimental evaluation of these algorithms also is provided. Our experiments indicate that for the fixed-slot problem, the minimum-hop feasible path algorithm proposed by us in [8] maximizes network utilization for large networks while the dynamic adaptive feasible path algorithm proposed in this paper does this for small networks.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117146454","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 wireless sensor network composes of sensors with limited battery power. And, consequently, to save sensor's energy effectively so as to extend the lifetime of the sensor network is one of important issues for the network deployment. In this article, we assume that limited sense nodes are equipped with high power and moving capabilities in the cluster based sensor network. The main objective of these high power nodes is to support the cluster header, which is going to exhaust its power, for data forwarding. The power consumption required for the movement is much larger than that of data forwarding. If the mobile node moves frequently, it will exhaust its power shortly. In this paper, a novel approach is proposed to effectively guide the movement of the high power nodes by considering the above issues. The simulation results show that the proposed approach can balance the sensor energy and extend the network lifetime when compared to the LEACH-C scheme for the same network environment.
{"title":"Study of Cluster-Based Data Forwarding in Sensor Networks with Limited High-Power Mobile Nodes","authors":"Yen-Wen Chen, Hwa-Chun Ma, I-Hsuan Peng, Ta-En Chang","doi":"10.1109/I-SPAN.2008.48","DOIUrl":"https://doi.org/10.1109/I-SPAN.2008.48","url":null,"abstract":"The wireless sensor network composes of sensors with limited battery power. And, consequently, to save sensor's energy effectively so as to extend the lifetime of the sensor network is one of important issues for the network deployment. In this article, we assume that limited sense nodes are equipped with high power and moving capabilities in the cluster based sensor network. The main objective of these high power nodes is to support the cluster header, which is going to exhaust its power, for data forwarding. The power consumption required for the movement is much larger than that of data forwarding. If the mobile node moves frequently, it will exhaust its power shortly. In this paper, a novel approach is proposed to effectively guide the movement of the high power nodes by considering the above issues. The simulation results show that the proposed approach can balance the sensor energy and extend the network lifetime when compared to the LEACH-C scheme for the same network environment.","PeriodicalId":305776,"journal":{"name":"2008 International Symposium on Parallel Architectures, Algorithms, and Networks (i-span 2008)","volume":"95 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126260602","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：