203 results found.
Year: 2015, Volume: 15, Number: 4
... 2015 Figure 3. Mean Average Precision (MAP) while varying the number of cluster centers. make the Fisher Kernel computationally feasible, we apply increasing the number ... approximately 36,000 annotated frames; performance of the algorithm: the number of clusters c, the PEVID-HD: consists of 21 video clips (16 seconds each ...
Terms matched: 2 - Score: 5 - 1,553k
Year: 2015, Volume: 15, Number: 4
... , A. Graur, "Dual tasking: gait and tremor in 173, 1991. [Online]. Available: Parkinson's disease – acquisition, processing and clustering", Rev. http://dx.doi.org/10.1007/BF01209064 Roum. Sci. Techn. – Électrotechn. et Énerg., 58, 3 ...
Terms matched: 1 - Score: 3 - 1,104k
Year: 2015, Volume: 15, Number: 4
... practical way of determining the state of the power system cluster groups in the system. The proposed method is ... , but systems. A hybrid Calinski-Harabasz criterion and k-means clustering algorithm is developed for the determination of the practical ... RPR basins by using the Jacobian of the reactive power clusters before the application of the clustering algorithm; and ...
Terms matched: 5 - Score: 486 - 1,493k
Year: 2015, Volume: 15, Number: 4
... . [Online]. Available: effectively used for complex numerical association rules http://dx.doi.org/10.1287/ijoc.1.3.190 mining, sequential patterns mining, and clustering rules [15] E. Atashpaz-Gargari and C. Lucas, 2007, "Imperialist competitive mining tasks of data mining. Parallel and distributed version algorithm: ...
Terms matched: 1 - Score: 3 - 1,807k
Year: 2015, Volume: 15, Number: 4
... The push-pull decision in the chunk- of developing disconnected islands. RTT ranking-based peer wise protocol is made on an individual chunk basis; chunk selection would form a cluster of proximal peers somewhat rarity is measured in terms of the number of a particular isolated from the rest of the network. This network chunk's occurrences in the ...
Terms matched: 1 - Score: 3 - 1,280k
Year: 2015, Volume: 15, Number: 3
... I. INTRODUCTION assigned dimensions. Then, a subspace cluster becomes The proliferation of different types of measurements and ... Volume 15, Number 3, 2015 Evaluation of Subspace Clustering Using Internal Validity Measures Mariusz OSZUST, Michał KOSTKA ... [6, 8- Subspace clustering techniques can create clusters which utilise 13]. They fall into external ...
Terms matched: 6 - Score: 735 - 1,251k
Year: 2015, Volume: 15, Number: 3
... works pay little attention to combining geocompution (HPC) cluster with large amount of CPUs, memory, hard ... data manager, which implements: workflow system over HPC clusters. The first and foremost, large-scale dataset storing ... and J. Li, "A new framework of cluster-based parallel spatial analysis algorithms, geographical data management, ...
Terms matched: 3 - Score: 51 - 1,662k
Year: 2015, Volume: 15, Number: 3
... 10.1109/HPDC.1999.805285 resource allocation change dynamically. The effectiveness [10] P. Dinda, D. O'Hallaron, "Host load prediction using linear models", Cluster Computing, 2000. [Online]. Available: of any applied algorithm depends on the unavoidable http://dx.doi.org/10.1023/A :1019048724544 ...
Terms matched: 1 - Score: 3 - 1,408k
Year: 2015, Volume: 15, Number: 2
... on a hidden Markov model 40 times faster than the computing power of USV Roadrunner cluster [16]. The classic one does. Based on the performance obtained ... Number 2, 2015 Optimizing the Forward Algorithm for Hidden Markov Model on IBM Roadrunner clusters Stefania – Iuliana SOIMAN1, Ionela RUSU2, Stefan-Gheorghe PENTIUC3 Stefan cel Mare University ...
Terms matched: 2 - Score: 47 - 983k
Year: 2015, Volume: 15, Number: 2
... applied to WBCD database (breast cancer database) to cluster data 0 ï‚£ i0 ï‚£ 1 (3) ... Computer Engineering Volume 15, Number 2, 2015 Data Clustering on Breast Cancer Data Using Firefly Algorithm with Golden ... 2 1 3 1 1 4 Figure 4. Determining clusters by using the best matrix.Figure 1018099 4 2 1 ...
Terms matched: 4 - Score: 66 - 905k