The data analysis involves the performance of different tasks, which can be performed by many different techniques and strategies. This research emphasizes the performance of the Data Clustering task using the artific...
详细信息
ISBN:
(纸本)1601320728
The data analysis involves the performance of different tasks, which can be performed by many different techniques and strategies. This research emphasizes the performance of the Data Clustering task using the artificial neural network Self-Organizing Maps as the main artifact. SOM is an Artificial Neural Network based in a competitive unsupervised learning, what implies in the training being entirely guided by the data and the neurons of the map compete among themselves. This neural network has the ability to form mappings that quantify the data, preserving its topology. This work introduces a new methodology of data clustering since SOM, that considers the topological map generated by him and the data topology in the clustering process, once considers the neighborhood of the input data.
Complex technical systems, such as mechatronic systems, can exploit networking as well as the computational power available today to achieve an automatic improvement of the technical system performance at run-time thr...
详细信息
A novel digital redesign of the analog model-reference-based decentralized adaptive controller is proposed for the sampled-data large scale system consisting of N interconnected linear subsystems with actuator faults....
详细信息
ISBN:
(纸本)9780889867468
A novel digital redesign of the analog model-reference-based decentralized adaptive controller is proposed for the sampled-data large scale system consisting of N interconnected linear subsystems with actuator faults. The adaptation of the analog controller gain is derived by using the model-reference adaptive control theory based on Lyapunov method. In this paper, it is shown that the sampled-data decentralized adaptive control system is theoretically possible to asymptotically track the desired output with a specified performance even when actuator faults occur. Then, a method of actuator fault recovery is proposed. With the estimated faults, one can use the proposed input compensation method to solve actuator faults. In this paper, we also introduce a prediction-based digital redesign method to develop the corresponding sampled-data controller for the sampled-data decentralized adaptive control systems.
Nanometer technology is gradually being applied after deep submicron technology due to the rapid progress of the VLSI fabrication process. Recently, system- on-a-chip (SoC) based products are gaining more advantages s...
详细信息
We introduce a general framework for string kernels. This framework can produce various types of kernels, including a number of existing kernels, to be used with support vector machines (SVMs). In this framework, we c...
详细信息
ISBN:
(纸本)9781848161085
We introduce a general framework for string kernels. This framework can produce various types of kernels, including a number of existing kernels, to be used with support vector machines (SVMs). In this framework, we can select the informative subsequences to reduce the dimensionality of the feature space. We can model the mutations in biological sequences. Finally, we combine contributions of subsequences in a weighted fashion to get the target kernel. In practical computation, we develop a novel tree structure, coupled with a traversal algorithm to speed up the computation. The experimental results on a benchmark SCOP data set show that the kernels produced by our framework outperform the existing spectrum kernels, in both efficiency and ROC50 scores.
In this article, we have concentrated on selecting the reliable path from source to destination in a mobile ad-hoc network (MANET) framework using Dempster-Shafer Theory (DST) of evidence. The belief and plausibility ...
详细信息
ISBN:
(纸本)9780769535135
In this article, we have concentrated on selecting the reliable path from source to destination in a mobile ad-hoc network (MANET) framework using Dempster-Shafer Theory (DST) of evidence. The belief and plausibility functions are used here for calculating the suitable path for sending data packet from source to destination selecting the proper neighborhood oil the basis of radio range of the source node. The radio range is determined depending on the Proximity Index and birth rate and death rate of the neighborhood nodes of the source node. Whenever data packet is changing hops within source and destination some credit and penalty are introduced to the next and previous hops respectively. In the concluding section of the present article a measure of gain (in delay) has been introduced for comparison of the existing methods vis-a-vis that of the proposed one.
In the paper, a new parallel LZW-Like algorithm, bidirectory LZW algorithm (BD-LZW) will be interpreted The new algorithm can be used in data compression/decompression system which runs on multi-microprocessor system,...
详细信息
The data analysis involves the performance of different tasks, which can be performed by many different techniques and strategies. This research emphasizes the performance of the Data Clustering task using the artific...
详细信息
ISBN:
(纸本)9781601320728
The data analysis involves the performance of different tasks, which can be performed by many different techniques and strategies. This research emphasizes the performance of the Data Clustering task using the artificial neural network Self-Organizing Maps as the main artifact. SOM is an Artificial Neural Network based in a competitive unsupervised learning, what implies in the training being entirely guided by the data and the neurons of the map compete among themselves. This neural network has the ability to form mappings that quantify the data, preserving its topology. This work introduces a new methodology of data clustering since SOM, that considers the topological map generated by him and the data topology in the clustering process, once considers the neighborhood of the input data.
The problem of multiple data sources selection (MDSS) in DSE (data-sharing environments) is addressed and the algorithm MDSSA (MDSS algorithm) is presented. MDSSA introduces the concept of Pareto optimization which re...
详细信息
The problem of multiple data sources selection (MDSS) in DSE (data-sharing environments) is addressed and the algorithm MDSSA (MDSS algorithm) is presented. MDSSA introduces the concept of Pareto optimization which reduces the search space greatly. By means of a novel normal-measure based nonlinear cost function, MDSSA computes approximate Pareto optimal paths to each data source first, and then gives the optimal data source and its corresponding path by comparing the cost of all candidate paths, resulting in finding more effective paths and much shorter response time. Extensive simulations show the efficiency of the algorithm.
This paper proposes a distributed energy efficient clustering protocol for target surveillance in sensor networks (EECTS). The protocol selects cluster heads according to a hybrid of node's residual energy and dis...
详细信息
This paper proposes a distributed energy efficient clustering protocol for target surveillance in sensor networks (EECTS). The protocol selects cluster heads according to a hybrid of node's residual energy and distribution of its neighbors. In addition, for the sake of reducing the energy dissipation of the cluster head, a minimum spanning tree with root of the base station is constructed among the cluster heads. Then it sends the gathered data to its upstream node along the spanning tree. The chief task of surveillance sensor networks is to sensing the moving target. So an intra-cluster node schedule method named EECTS-1 that can senses the most part of the network and it's enhanced method EECTS-2 are introduced. The two methods can obtain the high continuous surveillance degree when the moving target enters into the network. EECTS produces a linear network lifetime in the number of nodes and keeps good continuous surveillance degree simultaneously. The simulation results show that with the same performance of surveillance the EECTS-1 achieves the same lifetime as that HEED obtains and outperforms DEEG with up to 35% improvement. The EECTS-2 outperforms HEED significantly with prolonging the network lifetime about 70%-80%. Therefore the EECTS is suitable for military target surveillance and has high reliability of the sensing information.
暂无评论