Supervised feature selection determines feature relevance by evaluating feature's correlation with the classes. Joint minimization of a classifier's loss function and an 2;1-norm regularization has been shown ...
详细信息
ISBN:
(纸本)9781577356332
Supervised feature selection determines feature relevance by evaluating feature's correlation with the classes. Joint minimization of a classifier's loss function and an 2;1-norm regularization has been shown to be effective for feature selection. However, the appropriate feature subset learned from different classifiers' loss function may be different. Less effort has been made on improving the performance of feature selection by the ensemble of different classifiers' criteria and take advantages of them. Furthermore, for the cases when only a few labeled data per class are available, overfitting would be a potential problem and the performance of each classifier is restrained. In this paper, we add a joint 2;1-norm on multiple feature selection matrices to ensemble different classifiers' loss function into a joint optimization framework. This added co-regularization term has twofold role in enhancing the effect of regularization for each criterion and uncovering common irrelevant features. The problem of over-fitting can be alleviated and thus the performance of feature selection is improved. Extensive experiment on different data types demonstrates the effectiveness of our algorithm.
The composition of the landscape image can convey the emotion of cameramen. Proposed in this article is one novel technique for landscape image composition analysis, which could lay the foundation for the further anal...
详细信息
This paper proposes an adaptive video super-resolution (SR) method based on superpixel-guided auto-regressive (AR) model. The keyframes are automatically selected and super-resolved by a sparse regression method. The ...
详细信息
Tensors are increasingly common in several areas such as data mining, computer graphics, and computer vision. Tensor clustering is a fundamental tool for data analysis and pattern discovery. However, there usually exi...
详细信息
ISBN:
(纸本)9781577356332
Tensors are increasingly common in several areas such as data mining, computer graphics, and computer vision. Tensor clustering is a fundamental tool for data analysis and pattern discovery. However, there usually exist outlying data points in realworld datasets, which will reduce the performance of clustering. This motivates us to develop a tensor clustering algorithm that is robust to the outliers. In this paper, we propose an algorithm of Robust Tensor Clustering (RTC). The RTC firstly finds a lower rank approximation of the original tensor data using a L1 norm optimization function. Because the L1 norm doesn't exaggerate the effect of outliers compared with L2 norm, the minimization of the L1 norm approximation function makes RTC robust to outliers. Then we compute the HOSVD decomposition of this approximate tensor to obtain the final clustering results. Different from the traditional algorithm solving the approximation function with a greedy strategy, we utilize a non-greedy strategy to obtain a better solution. Experiments demonstrate that RTC has better performance than the state-ofthe- art algorithms and is more robust to outliers.
In spite of the fact that a fingerprint image may suffer from problems like noises and distortions, estimating core and delta points (singular points) is crucial for most of the automatic fingerprint identification sy...
详细信息
ISBN:
(纸本)9781467350891
In spite of the fact that a fingerprint image may suffer from problems like noises and distortions, estimating core and delta points (singular points) is crucial for most of the automatic fingerprint identification system (AFIS). Singular points have wide range of uses in AFIS which include fingerprint alignment and classification. This paper presents a simple method to locate singular points from fingerprint images using orientation consistency measure. Since consistency of the orientation field is minimum in singular regions, this method will find the local minimum to accurately detect the singular points. Core and delta points are distinguished by using the orientation field change. Experimental results show that our proposed algorithm can detect singular points effectively.
This paper offers a synthesized approach of solving the shortage of the traditional similarity in ontology mapping. First, it selects high correlation concepts by Hirst-St-Onge semantic relativity algorithms, in order...
详细信息
This paper presents a fully parallelized and scalable RNS Montgomery multiplier over binary *** generalizing the RNS Montgomery Multiplication (RNS MM) and pseudo-Mersenne-like numbers, we are able to obtain a conside...
详细信息
This paper presents a fully parallelized and scalable RNS Montgomery multiplier over binary *** generalizing the RNS Montgomery Multiplication (RNS MM) and pseudo-Mersenne-like numbers, we are able to obtain a considerably high speed in our FPGA implementation experiments with acceptable circuit area and modest critical path ***, this design can be easily scalable by adjusting a variety of field sizes and field polynomials.
Many practical problems can be converted to solving a polynomial equations in triangular form. For the sake of real-time control, the engineers want to simplify the polynomial equations usually, so that every polynomi...
详细信息
Dixon's method is a practical approach for computing the multivariate resultant in the quantifier elimination. The method is found to be quite restrictive as the Dixon matrix is singular always, and the Dixon resu...
详细信息
Learning to rank became a hot research topic in recent years and utilizing relational information in list-wise algorithms was discovered to be valuable and was widely adopted in various algorithms. These algorithms...
详细信息
暂无评论