Similarity computation is especially significant in collaborative filtering algorithms. In the existed literatures or large recommender systems, researchers generally use cosine similarity or Pearson correlation coeff...
详细信息
Being able to understand dynamics of human mobility is essential for social management and urban planning. Based on the sociology of community, this paper chooses different types of communities as research objects, an...
详细信息
Dialog Act (DA) is an important pragmatics feature for us to understand speakers' intention. Many methods have been proposed to recognize DA tags. However, little work has been conducted to address the problem of ...
详细信息
In recent years, the Web of Data has emerged with the release of growing amount of Linked Data. Since traditional Information Retrieval (IR) technologies are no longer suit for the retrieval on Linked Data, it becomes...
详细信息
Visual saliency detection and segmentation are widely used in many applications in image processing and computer vision. However, existing saliency detection methods have not fully taken the spatial information of sal...
详细信息
The long-term hemispheric variation of the flare index is investigated. It is found that, （1） the phase difference of the flare index between the northern and southern hemispheres is about 6-7 months, which is near ...
详细信息
The long-term hemispheric variation of the flare index is investigated. It is found that, （1） the phase difference of the flare index between the northern and southern hemispheres is about 6-7 months, which is near the time delay between flare activity and sunspot activity; （2） both the dominant and phase-leading hemisphere of the flare index is the northern hemisphere in the considered time interval, implying that the hemispheric asynchrony of solar activity has a close connection with the N-S asymmetry of solar activity.
The computational discovery of DNA motifs for previously uncharacterized transcription factors in groups of co-regulated genes is a well-studied problem with a great deal of practical relevance to the biologist. As a ...
详细信息
Online storage and sharing for large-scale DICOM images becomes increasingly important for medical organizations or large hospitals. This paper presents a distributed architecture based on Hadoop and HBase to support ...
详细信息
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.
In the scene classification, a scene can be considered as a set of object cliques. Objects inside each clique have semantic correlations with each other, while two objects from different cliques are relatively indepen...
详细信息
暂无评论