Unsupervised feature selection has become an important and challenging problem faced with vast amounts of unlabeled and high-dimension data in machine learning. We propose a novel unsupervised feature selection method...Unsupervised feature selection has become an important and challenging problem faced with vast amounts of unlabeled and high-dimension data in machine learning. We propose a novel unsupervised feature selection method using Structured Self-Representation( SSR) by simultaneously taking into account the selfrepresentation property and local geometrical structure of features. Concretely,according to the inherent selfrepresentation property of features,the most representative features can be selected. Mean while,to obtain more accurate results,we explore local geometrical structure to constrain the representation coefficients to be close to each other if the features are close to each other. Furthermore,an efficient algorithm is presented for optimizing the objective function. Finally,experiments on the synthetic dataset and six benchmark real-world datasets,including biomedical data,letter recognition digit data and face image data,demonstrate the encouraging performance of the proposed algorithm compared with state-of-the-art algorithms.展开更多
The formation of sintering necks between two metal fibers was investigated using the oval-oval model with respect to the fiber angle range of 0°-90°. Surface diffusion was assumed to be the predominant mecha...The formation of sintering necks between two metal fibers was investigated using the oval-oval model with respect to the fiber angle range of 0°-90°. Surface diffusion was assumed to be the predominant mechanism in every section of the junction of two metal fibers in this model, which was addressed numerically using the level- set method. The growth rates of the sintering necks in the direction of the bisector of obtuse angle, the bisector of acute angle and the fiber axis were discussed in detail. It is found that the growth rate of the sintering necks decreases with fiber angle increasing in the direction of the fiber axis and the bisector of acute angle. However, an opposite variation in growth rate of sintering necks can be found in the direction of the bisector of obtuse angle. The numerical simulation results show that the growth rate of the sintering necks is significantly affected by the initial local geomet- rical structure which is determined by the fiber angle.展开更多
基金Sponsored by the Major Program of National Natural Science Foundation of China(Grant No.13&ZD162)the Applied Basic Research Programs of China National Textile and Apparel Council(Grant No.J201509)
文摘Unsupervised feature selection has become an important and challenging problem faced with vast amounts of unlabeled and high-dimension data in machine learning. We propose a novel unsupervised feature selection method using Structured Self-Representation( SSR) by simultaneously taking into account the selfrepresentation property and local geometrical structure of features. Concretely,according to the inherent selfrepresentation property of features,the most representative features can be selected. Mean while,to obtain more accurate results,we explore local geometrical structure to constrain the representation coefficients to be close to each other if the features are close to each other. Furthermore,an efficient algorithm is presented for optimizing the objective function. Finally,experiments on the synthetic dataset and six benchmark real-world datasets,including biomedical data,letter recognition digit data and face image data,demonstrate the encouraging performance of the proposed algorithm compared with state-of-the-art algorithms.
基金financially supported by the National Natural Science Foundation of China (Nos. 51174236 and 51134003)the National Basic Research Program of China (No. 2011CB606306)the Opening Project of State Key Laboratory of Porous Metal Materials (No. PMM-SKL-4-2012)
文摘The formation of sintering necks between two metal fibers was investigated using the oval-oval model with respect to the fiber angle range of 0°-90°. Surface diffusion was assumed to be the predominant mechanism in every section of the junction of two metal fibers in this model, which was addressed numerically using the level- set method. The growth rates of the sintering necks in the direction of the bisector of obtuse angle, the bisector of acute angle and the fiber axis were discussed in detail. It is found that the growth rate of the sintering necks decreases with fiber angle increasing in the direction of the fiber axis and the bisector of acute angle. However, an opposite variation in growth rate of sintering necks can be found in the direction of the bisector of obtuse angle. The numerical simulation results show that the growth rate of the sintering necks is significantly affected by the initial local geomet- rical structure which is determined by the fiber angle.
