Clustering is the process of gathering objects into groups based on their feature's similarity. In this paper, we concentrate on Weighted Kernel K-Means method for its capability to manage nonlinear separability a...
详细信息
Clustering is the process of gathering objects into groups based on their feature's similarity. In this paper, we concentrate on Weighted Kernel K-Means method for its capability to manage nonlinear separability and high dimensionality in the data. A new slight modification of WKM algorithm has been proposed and tested on real Rice data. The results show that the accuracy of proposed algorithm is higher than other famous clustering algorithm and ensures that the WKM is a good solution for real world problems.
In this paper, the cross-entropy (CE) method is proposed to solve non-linear discriminant analysis or kernel Fisher discriminant (CE-KFD) analysis. CE through certain steps can find the optimal or near optimal solutio...
详细信息
In this paper, the cross-entropy (CE) method is proposed to solve non-linear discriminant analysis or kernel Fisher discriminant (CE-KFD) analysis. CE through certain steps can find the optimal or near optimal solution with a fast rate of convergence for optimization problem. While, KFD is to solve problem of Fisher's linear discriminant in a kernel feature space F by maximizing between class variance and minimizing within class variance. Through the numerical experiments, we found that CE-KFD demonstrates the high accuracy of the results compared to the traditional methods, Fisher LDA and kernel Fisher (KFD) with eigen decomposition method.
This paper proposes a new approach based on 4×4 discrete orthogonal Tchebichef moment for fast and efficient image compression. The method incorporates a simplified mathematical framework technique using matrices...
详细信息
This paper proposes a new approach based on 4×4 discrete orthogonal Tchebichef moment for fast and efficient image compression. The method incorporates a simplified mathematical framework technique using matrices, as well as a block-wise reconstruction technique to eliminate possible occurrences of numerical instabilities at higher moment orders. The comparison between 4×4 Tchebichef moment transform and discrete cosine transform has been done. The results show significant advantages for 4×4 Tchebichef moment in terms of its error reconstruction and average bit-length of Huffman codes.
To segment foreground objects and moving shadows in visual surveillance environment, this paper proposes an algorithm by exploiting color information, illumination invariants and spatial information. The presence of a...
详细信息
ISBN:
(纸本)9780769536422
To segment foreground objects and moving shadows in visual surveillance environment, this paper proposes an algorithm by exploiting color information, illumination invariants and spatial information. The presence of a shadow is first hypothesized with simple evidence that shadows darken the surface which they are cast upon. Derivatives of illumination invariants are then used to classify the potential shadow pixels extracted in previous step. To increase the accuracy of shadow detection, two types of spatial analysis are designed to verify actual shadow pixels. Experimental results show that the proposed algorithm can detect moving shadow effectively on indoor and outdoor video sequences. The performance of the method is considerably higher than that of the two well-known shadow detection methods, and it is robust against changing illumination.
Many cerebellar learning theories assume that;long-term depression (LTD) of synapses between parallel fibres (PFs) and Purkinje cells (PCs) provides the basis for patternrecognition in the cerebellum. Previous work h...
详细信息
ISBN:
(纸本)9783642049200
Many cerebellar learning theories assume that;long-term depression (LTD) of synapses between parallel fibres (PFs) and Purkinje cells (PCs) provides the basis for patternrecognition in the cerebellum. Previous work has suggested that PCs can use a novel neural code based on the duration of silent periods. These simulations have used a simplified learning rule, where the synaptic conductance was halved each time a pattern was learned. However, experimental studies in cerebellar slices show that the synaptic conductance saturates and is rarely reduced to less than 50% of its baseline value. Moreover, the previous simulations did not include plasticity of the synapses between inhibitory interneurons and PCs. Here we study the effect of LTD saturation and inhibitory synaptic plasticity on patternrecognition in a complex PC model. We find that the PC model is very sensitive to the value at which LTD saturates, but is unaffected by inhibitory synaptic plasticity.
Local link analysis of topical graphs on the Web allows to experiment focused crawling strategies in a detailed way. In this scope, models, parameters and metrics used to orientate the crawler can be better understood...
详细信息
Local link analysis of topical graphs on the Web allows to experiment focused crawling strategies in a detailed way. In this scope, models, parameters and metrics used to orientate the crawler can be better understood, tuned and evaluated. We develop a methodological and experimental approach that exploits link analysis in order to determine what constitutes a good content analysis metric able to guide efficiently topical crawlers toward highly relevant areas of the Web. Our experimentations show that partial knowledge of the local topology of topical graph highlights our understanding of routing capabilities of various metrics. Furthermore, our experimentations demonstrate that significant crawling efficiency improvement can be reached.
An intelligent control chart patternrecognition system is essential for efficient monitoring and diagnosis process variation in automated manufacturing environment. Artificial neural networks (ANN) have been applied ...
详细信息
An intelligent control chart patternrecognition system is essential for efficient monitoring and diagnosis process variation in automated manufacturing environment. Artificial neural networks (ANN) have been applied for automated recognition of control chart patterns since the last 20 years. In early study, the development of control chart patterns recognizers was mainly based on generalized-ANN model. There has been an increasing trend among researchers to move beyond generalized recognizer particularly for addressing complex recognition tasks. However, the existing works mainly focus on univariate process cases. This paper aims to investigate an effective synergistic-ANN model for on-line monitoring and diagnosis multivariate process patterns. The recognition performances of a generalized-ANN and the parallel distributed ANN recognizers for learning dynamic patterns of multivariate process patterns were discussed.
This work presents study and experimentation for object recognition when isolated objects are under discussion. The circumstances of similarity transformations, presence of noise, and occlusion have been included as t...
详细信息
This work presents study and experimentation for object recognition when isolated objects are under discussion. The circumstances of similarity transformations, presence of noise, and occlusion have been included as the part of the study. For simplicity, instead of objects, outlines of the objects have been used for the whole process of the recognition. Fourier descriptors have been used as features of the objects. From the analysis and results using Fourier descriptors, the following questions arise: What is the optimum number of descriptors to be used? Are these descriptors of equal importance? To answer these questions, the problem of selecting the best descriptors has been formulated as an optimization problem. Genetic algorithm technique has been mapped and used successfully to have an object recognition system using minimal number of Fourier descriptors. The proposed method assigns, for each of these descriptors, a weighting factor that reflects the relative importance of that descriptor.
Recently, fusion speed has emerged as an important factor in the image fusion and a substantial amount of memory and computing power are required for a high-speed fusion. This paper shows approaches to accelerate mult...
详细信息
Recently, fusion speed has emerged as an important factor in the image fusion and a substantial amount of memory and computing power are required for a high-speed fusion. This paper shows approaches to accelerate multi-scale image fusion speed on GPU (graphics processing unit) using CUDA (compute unified device architecture). The GPU has evolved into a very powerful and flexible streaming processor, which provides a good computational power and memory bandwidth. We implement the multi-scale image fusion algorithms using CUDA software platform of the latest version of GPU and theirs fusion speeds are compared and evaluated with implementation in Core2 Quad processor with 2.4 GHz. The GPU version achieved a speedup of 6x-8x over the CPU version.
The bacterial foraging optimization (BFO) algorithm is a nature and biologically inspired computing method. We propose an alternative solution integrating bacterial foraging optimization algorithm and tabu search (TS)...
详细信息
The bacterial foraging optimization (BFO) algorithm is a nature and biologically inspired computing method. We propose an alternative solution integrating bacterial foraging optimization algorithm and tabu search (TS) algorithm namely TS-BFO. We modify the original BFO via established a self-control multi-length chemotactic step mechanism, and introduce rao metric. We utilize it to solve motif discovery problem and compare the experimental result with existing famous DE/EDA algorithm which combines global information extracted by estimation of distribution algorithm (EDA) with differential information obtained by Differential evolution (DE) to search promising solutions. The experiments on real data set selected from TRANSFAC and SCPD database have predicted meaningful motif which demonstrated that TS-BFO and DE/EDA are promising approaches for finding motif and enrich the technique of motif discovery.
暂无评论