This paper is to discuss the reduction of computation complexity in decision tree generation for the numerical-valued attributes. The proposed method is based on the partition impurity. The partition impurity minimiza...
详细信息
This paper is to discuss the reduction of computation complexity in decision tree generation for the numerical-valued attributes. The proposed method is based on the partition impurity. The partition impurity minimization is used to select the expanded attribute for generation the sub-node during the tree growth. After inducing the unstable cut-points of numerical-attributes, it is analytically proved that the partition impurity minimization can always be obtained at the unstable cut-points. It implies that the computation on stable cut-points may not be considered during the tree growth. Since the stable cut-points are far more than unstable cut-points, the experimental results show that the proposed method can reduce the computational complexity greatly.
Classification is necessary and basic to scientific research. The Chinese loanword has always been a hot spot of studies on Chinese linguistics, however the range of it remained unsettled. The paper will introduce fuz...
详细信息
Classification is necessary and basic to scientific research. The Chinese loanword has always been a hot spot of studies on Chinese linguistics, however the range of it remained unsettled. The paper will introduce fuzzy set technology into the discriminant process of Chinese loanwords to compose a reliable and efficient classifier. Simulations verify the efficiency and feasibility.
During recent years, there are more and more high-quality information in the Web database. Thus, it is becoming more and more important to find the most relevant Web database to user's query. In this paper, we pro...
详细信息
A new method, multi-class fuzzy support vector machine of dismissing margin (DMFSVM) based on class-center, is proposed aiming at the outliers and noises which appear in the large quantity samples. Compared with tradi...
详细信息
A new method, multi-class fuzzy support vector machine of dismissing margin (DMFSVM) based on class-center, is proposed aiming at the outliers and noises which appear in the large quantity samples. Compared with traditional SVM, this new method eliminates the sensibility of optimal separating hyperplane. It weeds out some sample points which may not be support vectors. And this will be able to decrease the corresponding optimization problem dimension, reduce the memory and the amount of computation, but increase the training speed. At the same time, the new algorithm adopts fuzzy membership function of decreasing Semi-Cauchy type. The advantages are through regulating parameters of fuzzy factor suitably according to the specific circumstances to make the fuzzy factor of isolated points smaller and the fuzzy factor of support vectors larger relatively. So this method can fit the characteristics of fuzzy classification well.
In this study, we study set operations on type-2 fuzzy sets. We first discuss join and meet operations of membership grades of type-2 fuzzy sets under left continuous t-norms and derive distributive law of type-2 fuzz...
详细信息
In this study, we study set operations on type-2 fuzzy sets. We first discuss join and meet operations of membership grades of type-2 fuzzy sets under left continuous t-norms and derive distributive law of type-2 fuzzy sets. Then, some properties on compositions of fuzzy relations is discussed. We derived that the distributive laws under union and composition of type-2 fuzzy relations is valid. An example shows the failure of distributive laws under intersection and composition.
A new method to solve the convex hull problem in n-dimensional spaces is proposed in this paper. At each step, a new point is added into the convex hull if the point is judged to be out of the current convex hull by a...
详细信息
A new method to solve the convex hull problem in n-dimensional spaces is proposed in this paper. At each step, a new point is added into the convex hull if the point is judged to be out of the current convex hull by a linear programming model. For the linear separable classification problem, if an instance is regarded as a point of the instances space, the overlap does not still occur between the convex hulls of different classes after a feature is deleted, then we can delete that feature. Repeat this process, an algorithm for feature selection is given. Experimental results show the effectiveness of the algorithm.
MCS (Minimal Consistent Set) is one of the classical algorithms for minimal consistent subset selection problem. However, when noisy samples are present classification accuracy can suffer. In addition, noise affect th...
详细信息
MCS (Minimal Consistent Set) is one of the classical algorithms for minimal consistent subset selection problem. However, when noisy samples are present classification accuracy can suffer. In addition, noise affect the size of minimal consistent set. Therefore, removing noise is an important issue before sample selection. In this paper, an improvement approach based on MCS to select the representative samples is proposed. Compared with other algorithms which remove the noise by Wilson Editing in advance for the representative samples selection, this algorithm performs the processes of noise removing and samples selection simultaneously. According to this method, most noise can be deleted and the most representative samples can be identified and retained. The experiments show that the proposed method can greatly remove the redundant samples and noise as well as increase the accuracy of solutions when it is used for classification tasks.
This paper presents an approach to instance selection for the nearest neighbor rule which aims to obtain a condensed set with high condensing rate and prediction accuracy. By making an improvement on MCS algorithm and...
详细信息
This paper presents an approach to instance selection for the nearest neighbor rule which aims to obtain a condensed set with high condensing rate and prediction accuracy. By making an improvement on MCS algorithm and allowing certain error rate on the training set, a condensed set with high condensing rate and satisfying prediction accuracy is obtained. The condensed set is order-independent of the training instances and insensitive to noise. Comparative experiments have been conducted on real data sets, and the results show its superiority to MCS and FCNN in terms of condensing rate and prediction accuracy.
There may be many fuzzy attributes in a fuzzy information system. Different fuzzy attribute has different contribution to classification. More important attributes have more contribution than the others to decision-ma...
详细信息
There may be many fuzzy attributes in a fuzzy information system. Different fuzzy attribute has different contribution to classification. More important attributes have more contribution than the others to decision-making. In this paper, based on the importance of the fuzzy condition attributes, a new method generating a fuzzy decision tree is proposed, which uses the important degree of the fuzzy condition attribute with respect to the fuzzy decision attributes to select attributes to expand the branches of a fuzzy decision tree. A comparison between the new method and fuzzy ID3 is provided. It is shown that the new method is more efficient than fuzzy ID3.
This paper proposes an image recognition method, which consists of two steps: features extraction based on wavelet transform and image recognition using artificial neural networks. More specifically, wavelet transform...
详细信息
This paper proposes an image recognition method, which consists of two steps: features extraction based on wavelet transform and image recognition using artificial neural networks. More specifically, wavelet transform is used to decompose the original image into different frequency sub-bands, then a set of features are extracted from the wavelet coefficients. The feature set as input fed into neural network for recognition. The experimental results confirmed effectiveness of the proposed approach.
暂无评论