Relative Density

14,000,000 Leading Edge Experts on the ideXlab platform

Scan Science and Technology

Contact Leading Edge Experts & Companies

Scan Science and Technology

Contact Leading Edge Experts & Companies

The Experts below are selected from a list of 285888 Experts worldwide ranked by ideXlab platform

Haitao Zou - One of the best experts on this subject based on the ideXlab platform.

  • Fuzzy Support Vector Machine With Relative Density Information for Classifying Imbalanced Data
    IEEE Transactions on Fuzzy Systems, 2019
    Co-Authors: Changyin Sun, Xibei Yang, Shang Zheng, Haitao Zou
    Abstract:

    Fuzzy support vector machine (FSVM) has been combined with class imbalance learning (CIL) strategies to address the problem of classifying skewed data. However, the existing approaches hold several inherent drawbacks, causing the inaccurate prior data distribution estimation, further decreasing the quality of the classification model. To solve this problem, we present a more robust prior data distribution information extraction method named Relative Density, and two novel FSVM-CIL algorithms based on the Relative Density information in this paper. In our proposed algorithms, a K-nearest neighbors-based probability Density estimation (KNN-PDE) alike strategy is utilized to calculate the Relative Density of each training instance. In particular, the Relative Density is irrelevant with the dimensionality of data distribution in feature space, but only reflects the significance of each instance within its class; hence, it is more robust than the absolute distance information. In addition, the Relative Density can better seize the prior data distribution information, no matter the data distribution is easy or complex. Even for the data with small injunctions or a large class overlap, the Relative Density information can reflect its details well. We evaluated the proposed algorithms on an amount of synthetic and real-world imbalanced datasets. The results show that our proposed algorithms obviously outperform to some previous work, especially on those datasets with sophisticated distributions.

Liu Qing - One of the best experts on this subject based on the ideXlab platform.

  • Relative Density-based Clustering Algorithm
    Computer Science, 2007
    Co-Authors: Liu Qing
    Abstract:

    With strong ability of discovery arbitrary shape clusters and handling noise,Density based clustering is one of primary methods for data mining.This paper provides a clustering algorithm based on Relative Density,which efficiently resolves these problem of being very sensitive to the user-defined parameters and too difficult for users to determine the parameters.

Changyin Sun - One of the best experts on this subject based on the ideXlab platform.

  • Fuzzy Support Vector Machine With Relative Density Information for Classifying Imbalanced Data
    IEEE Transactions on Fuzzy Systems, 2019
    Co-Authors: Changyin Sun, Xibei Yang, Shang Zheng, Haitao Zou
    Abstract:

    Fuzzy support vector machine (FSVM) has been combined with class imbalance learning (CIL) strategies to address the problem of classifying skewed data. However, the existing approaches hold several inherent drawbacks, causing the inaccurate prior data distribution estimation, further decreasing the quality of the classification model. To solve this problem, we present a more robust prior data distribution information extraction method named Relative Density, and two novel FSVM-CIL algorithms based on the Relative Density information in this paper. In our proposed algorithms, a K-nearest neighbors-based probability Density estimation (KNN-PDE) alike strategy is utilized to calculate the Relative Density of each training instance. In particular, the Relative Density is irrelevant with the dimensionality of data distribution in feature space, but only reflects the significance of each instance within its class; hence, it is more robust than the absolute distance information. In addition, the Relative Density can better seize the prior data distribution information, no matter the data distribution is easy or complex. Even for the data with small injunctions or a large class overlap, the Relative Density information can reflect its details well. We evaluated the proposed algorithms on an amount of synthetic and real-world imbalanced datasets. The results show that our proposed algorithms obviously outperform to some previous work, especially on those datasets with sophisticated distributions.

Kenji Ishihara - One of the best experts on this subject based on the ideXlab platform.

  • empirical correlation between spt n value and Relative Density for sandy soils
    Soils and Foundations, 1999
    Co-Authors: M Cubrinovski, Kenji Ishihara
    Abstract:

    ABSTRACT It has been known that the penetration resistance depends on the grain size of soils, and that fines-containing sands have smaller SPT N-values than clean sands. Previous investigations on the link between the penetration resistance and Relative Density have indicated that the ratio between the normalized N-value and the square of the Relative Density, N1/D2r, is dependent on the grain size of sands. While this dependence has not been properly quantified, it has been customary to utilize the well-known expression of Meyerhof with N1/D2r being fixed at 41. In this paper, an attempt was made to correlate the penetration resistance and Relative Density by accounting for the grain size properties of soils. The void ratio range (emax – emin) was used as a measure indicative of the grain size and grain size composition. It was found that M is highly dependent on the value of (emax – emin) and that this ratio gradually decreases with increasing void ratio range from a value of about 100 for gravels to a value of about 10 for silty sands. This dependence is mathematically formulated and used to establish an empirical correlation between the SPT TV-value and Dr that is applicable to various kinds of soils ranging from silty sands to gravels. The correlation is developed by using data of high-quality undisturbed samples and results of SPT measurements on natural deposits of sandy soils and gravels.

Ma De-liang - One of the best experts on this subject based on the ideXlab platform.

  • Data Stream Fuzzy Clustering Algorithm Based on Relative Density
    Computer Science, 2010
    Co-Authors: Ma De-liang
    Abstract:

    This paper provided a Relative Density based data stream fuzzy clustering algorithm which inherits the advantages of Relative Density based clustering and fuzzy clustering,so it can discover arbitrary-shape and multi-resolution clusters.With the subtraction operator on the set of micro-clusters which is defined according to the spatial overlapping relations among micro-clusters,this algorithm can do clustering on any user-specified data stream window.Compared with CluStream algorithm on the two areas of clustering quality and processing time,this algorithm demonstrates a clear advantage.