In order to select effective feature subsets for pattern classification, a novel statistics rough set method is presented based on generalized attribute reduction. Unlike classical reduction approaches, the objects in...In order to select effective feature subsets for pattern classification, a novel statistics rough set method is presented based on generalized attribute reduction. Unlike classical reduction approaches, the objects in universe of discourse are signs of training sample sets and values of attributes are taken as statistical parameters. The binary relation and discernibility matrix for the reduction are induced by distance function. Furthermore, based on the monotony of the distance function defined by Mahalanobis distance, the effective feature subsets are obtained as generalized attribute reducts. Experiment result shows that the classification performance can be improved by using the selected feature subsets.展开更多
The Mahalanobis distance features proposed by P.C.Mahalanobis, an Indian statistician, can be used in an automatic on-line cutting tool condition monitoring process based on digital image processing. In this paper, a ...The Mahalanobis distance features proposed by P.C.Mahalanobis, an Indian statistician, can be used in an automatic on-line cutting tool condition monitoring process based on digital image processing. In this paper, a new method of obtaining Mahalanobis distance features from a tool image is proposed. The key of calculating Mahalanobis distance is appropriately dividing the object into several component sets. Firstly, a technique is proposed that can automatically divide the component groups for calculating Mahalanobis distance based on the gray level of wearing or breakage regions in a tool image. The wearing region can be divided into high gray level component group and the tool-blade into low one. Then, the relation between Mahalanobis distance features of component groups and tool conditions is investigated. The results indicate that the high brightness region on the flank surface of the turning tool will change with its abrasion change and if the tool is heavily abraded, the area of high brightness will increase apparently. The Mahalanobis distance features of high gray level component group are related with wearing state of tool and low gray level component group correlated with breakage of tool. The experimental results show that the abrasion of the tool’s flank surface affected the Mahalanobis distances of high brightness component of the tool and the pixels of high brightness component set. Compared with the changes of them, we found that the Mahalanobis distance of high brightness component of the tool was more sensitive to the abrasion of cutting tool than the area of high brightness component set of the tool. Here we found that the relative changing rate of the area of high brightness component set was not quite obvious and it was ranging from 2% to 15%, while the relative changing rate of the Mahalanobis distance in table 1 ranges from 13.9% to 47%. It is 3 times higher than the changing rate of the area.展开更多
Classical Mahalanobis distance is used as a method of detecting outliers, and is affected by outliers. Some robust Mahalanobis distance is proposed via the fast MCD estimator. However, the bias of the MCD estimator in...Classical Mahalanobis distance is used as a method of detecting outliers, and is affected by outliers. Some robust Mahalanobis distance is proposed via the fast MCD estimator. However, the bias of the MCD estimator increases significantly as the dimension increases. In this paper, we propose the improved Mahalanobis distance based on a more robust Rocke estimator under high-dimensional data. The results of numerical simulation and empirical analysis show that our proposed method can better detect the outliers in the data than the above two methods when there are outliers in the data and the dimensions of data are very high.展开更多
基金This work was supported by the National Basic Research Program of China(No.2001CB309403)
文摘In order to select effective feature subsets for pattern classification, a novel statistics rough set method is presented based on generalized attribute reduction. Unlike classical reduction approaches, the objects in universe of discourse are signs of training sample sets and values of attributes are taken as statistical parameters. The binary relation and discernibility matrix for the reduction are induced by distance function. Furthermore, based on the monotony of the distance function defined by Mahalanobis distance, the effective feature subsets are obtained as generalized attribute reducts. Experiment result shows that the classification performance can be improved by using the selected feature subsets.
文摘The Mahalanobis distance features proposed by P.C.Mahalanobis, an Indian statistician, can be used in an automatic on-line cutting tool condition monitoring process based on digital image processing. In this paper, a new method of obtaining Mahalanobis distance features from a tool image is proposed. The key of calculating Mahalanobis distance is appropriately dividing the object into several component sets. Firstly, a technique is proposed that can automatically divide the component groups for calculating Mahalanobis distance based on the gray level of wearing or breakage regions in a tool image. The wearing region can be divided into high gray level component group and the tool-blade into low one. Then, the relation between Mahalanobis distance features of component groups and tool conditions is investigated. The results indicate that the high brightness region on the flank surface of the turning tool will change with its abrasion change and if the tool is heavily abraded, the area of high brightness will increase apparently. The Mahalanobis distance features of high gray level component group are related with wearing state of tool and low gray level component group correlated with breakage of tool. The experimental results show that the abrasion of the tool’s flank surface affected the Mahalanobis distances of high brightness component of the tool and the pixels of high brightness component set. Compared with the changes of them, we found that the Mahalanobis distance of high brightness component of the tool was more sensitive to the abrasion of cutting tool than the area of high brightness component set of the tool. Here we found that the relative changing rate of the area of high brightness component set was not quite obvious and it was ranging from 2% to 15%, while the relative changing rate of the Mahalanobis distance in table 1 ranges from 13.9% to 47%. It is 3 times higher than the changing rate of the area.
文摘Classical Mahalanobis distance is used as a method of detecting outliers, and is affected by outliers. Some robust Mahalanobis distance is proposed via the fast MCD estimator. However, the bias of the MCD estimator increases significantly as the dimension increases. In this paper, we propose the improved Mahalanobis distance based on a more robust Rocke estimator under high-dimensional data. The results of numerical simulation and empirical analysis show that our proposed method can better detect the outliers in the data than the above two methods when there are outliers in the data and the dimensions of data are very high.
