In this paper, we investigate one-class and clustering problems by using statistical learning theory. To establish a universal framework, a unsupervised learning problem with predefined threshold η is formally descri...
详细信息
The process of nondominated sorting is one of main time-consuming parts of multiobjective evolutionary algorithm (MOEA). Designing a fast nondominated sorting algorithm is crucial to improve the performance of MOEA. T...
详细信息
The key issue of Peer Data Management Systems (PDMSs) is how to efficiently organize and manage distributed resources in P2P networks to accurately route queries from the peer initiating the query to appropriate peers...
详细信息
Analysis of transforming matrices between Bezier basis functions and geometrically continuous basis functions is presented It is shown that G 2 transforming matrix has some relationship with G1 transforming matrix. Ba...
详细信息
Feature extraction or selection is one of the most importmant steps in pattern recognition or pattern classification, data mining, machine learning and so on. In this paper, we introduce the information theory, propos...
详细信息
Principal component analysis (PCA) is an important method in multivariate statistical analysis, and its main idea is compression of dimensionality including variables and samples. In this paper, based on the ideas con...
详细信息
Emotion study is a multi-disciplinary research subject. In the past three decades, a number of theoretical models of emotion and computer applications have been proposed from different perspectives including psycholog...
详细信息
Based on Jordan curve theorem, a universal classification method based on hyper surface is recently put forward. The experiments show that the new method can efficiently and accurately classify large data size up to 1...
详细信息
Based on Jordan curve theorem, a universal classification method based on hyper surface is recently put forward. The experiments show that the new method can efficiently and accurately classify large data size up to 10/sup 7/ in three-dimensional space. However, the number of training samples needed to design a classifier grows with the dimension of the features. So a way to reduce the dimension of the features without losing any essential information is needed. We put forward a kind of simple and efficient dimension reduction method without losing any essential information to improve the performance of classification based on hyper surface for high dimension data.
In this paper, we present a brief summary to 3D mesh model segmentation techniques, including definition, latest achievements, classification and application in this field. Then evaluations on some of typical methods,...
详细信息
In this paper, we present a brief summary to 3D mesh model segmentation techniques, including definition, latest achievements, classification and application in this field. Then evaluations on some of typical methods, such as Watershed, topological and geometrical method, are introduced. After some applications are presented, problems and prospect of the techniques are also discussed.
This paper presents a novel segmentation method based on a non-parametric background model that has the ability of modeling multi-model. Firstly, both the intensity and edge features are used to improve robustness of ...
详细信息
This paper presents a novel segmentation method based on a non-parametric background model that has the ability of modeling multi-model. Firstly, both the intensity and edge features are used to improve robustness of the foreground detection. Secondly, we also present an adaptive shadow detection model to find the accurate moving objects. The experiment results show that our proposed method is effective.
暂无评论