Gaussian clustering method based on maximum-fuzzy-entropy interpretation

doi:10.1016/S0165-0114(97)00126-7

Fuzzy Sets and Systems

Volume 102, Issue 2, 1 March 1999, Pages 253-258

https://doi.org/10.1016/S0165-0114(97)00126-7 Get rights and content

Abstract

A new method of fuzzy clustering is proposed. This is a complete Gaussian membership function derived by means of the maximum-entropy interpretation. Compared to the traditional fuzzy c-means (FCM) method, our approach exhibits the following two advantages: (1) having clearer physical meaning and well-defined mathematical features; (2) having an optimal choice for feature parameter σ in theory. Moreover, we also review some existing measures of uncertainty of fuzzy sets, and redefine fuzzy entropy as analogous to probabilistic entropy.

References (13)

A. De Luca et al.
A definition of a nonprobabilistic entropy in the setting of fuzzy sets theory
Inform. Control
(1972)
B. Kosko
Neural Networks and Fuzzy Systems
(1991)
J. Aczel et al.
On Measures of Information and their Characterizations
(1975)
J.C. Bezdek
Pattern Recognition with Fuzzy Objective Function Algorithms
(1981)
S. Geman et al.
Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of image
IEEE Trans. Pattern Anal. Machine Intell.
(1984)
E.T. Jaynes
Information theory and statistical mechanics
Phys. Rev.
(1957)
E.T. Jaynes
Information theory and statistical mechanics
Phys. Rev.
(1957)

There are more references available in the full text version of this article.

Cited by (66)

Local and NON-LOCAL source apportionment of black carbon and combustion generated PM<inf>2.5</inf>
2024, Environmental Pollution
Current methods for measuring black carbon aerosol (BC) by optical methods apportion BC to fossil fuel and wood combustion. However, these results are aggregated: local and non-local combustion sources are lumped together. The spatial apportioning of carbonaceous aerosol sources is challenging in remote or suburban areas because non-local sources may be significant. Air quality modeling would require highly accurate emission inventories and unbiased dispersion models to quantify such apportionment. We propose FUSTA (FUzzy SpatioTemporal Apportionment) methodology for analyzing aethalometer results for equivalent black carbon coming from fossil fuel (eBC_ff) and wood combustion (eBC_wb).
We applied this methodology to ambient measurements at three suburban sites around Santiago, Chile, in the winter season 2021. FUSTA results showed that local sources contributed ∼80% to eBC_ff and eBC_wb in all sites. By using PM_2.5 – eBC_ff and PM_2.5 – eBC_wb scatterplots for each fuzzy cluster (or source) found by FUSTA, the estimated lower edge lines showed distinctive slopes in each measurement site. These slopes were larger for non-local sources (aged aerosols) than for local ones (fresh emissions) and were used to apportion combustion PM_2.5 in each site. In sites Colina, Melipilla and San Jose de Maipo, fossil fuel combustion contributions to PM_2.5 were 26 % (15.9 μg m⁻³), 22 % (9.9 μg m⁻³), and 22 % (7.8 μg m⁻³), respectively. Wood burning contributions to PM_2.5 were 22 % (13.4 μg m⁻³), 19 % (8.9 μg m⁻³) and 22% (7.3 μg m⁻³), respectively.
This methodology generates a joint source apportionment of eBC and PM_2.5, which is consistent with available chemical speciation data for PM_2.5 in Santiago.
Fuzzy k-Means: history and applications
2024, Econometrics and Statistics
The fuzzy approach to clustering arises to cope with situations where objects have not a clear assignment. Unlike the hard/standard approach where each object can only belong to exactly one cluster, in a fuzzy setting, the assignment is soft; that is, each object is assigned to all clusters with certain membership degrees varying in the unit interval. The best known fuzzy clustering algorithm is the fuzzy $k$ -means (F $k$ M), or fuzzy $c$ -means. It is a generalization of the classical $k$ -means method. Starting from the F $k$ M algorithm, and in more than 40 years, several variants have been proposed. The peculiarity of such different proposals depends on the type of data to deal with, and on the cluster shape. The aim is to show fuzzy clustering alternatives to manage different kinds of data, ranging from numeric, categorical or mixed data to more complex data structures, such as interval-valued, fuzzy-valued or functional data, together with some robust methods. Furthermore, the case of two-mode clustering is illustrated in a fuzzy setting.
A new methodology for source apportionment of gaseous industrial emissions
2023, Journal of Hazardous Materials
Air quality modeling (AQM) is often used to investigate gaseous pollution around industrial zones. However, this methodology requires accurate emission inventories, unbiased AQM algorithms and realistic boundary conditions.
We introduce a new methodology for source apportionment of industrial gaseous emissions, which is based on a fuzzy clustering of ambient concentrations, along with a standard AQM approach. First, by applying fuzzy clustering, ambient concentration is expressed as a sum of non-negative contributions — each corresponding to a specific spatiotemporal pattern (STP); we denote this method as FUSTA (FUzzy SpatioTemporal Apportionment). Second, AQM of the major industrial emissions in the study zone generates another set of STP. By comparing both STP sets, all major source contributions resolved by FUSTA are identified, so a source apportionment is achieved. The uncertainty in FUSTA results may be estimated by comparing results for different numbers of clusters.
We have applied FUSTA in an industrial zone in central Chile, obtaining the contributions from major sources of ambient SO₂: a thermal power plant complex and a copper smelter, and other contributions from local and regional sources (outside the AQM domain). The methodology also identifies SO₂ episodes associated to emissions from the copper smelter.
Fuzzy clustering algorithms with distance metric learning and entropy regularization
2021, Applied Soft Computing
Citation Excerpt :
□ Note that the proposed maximum entropy clustering algorithms share some similarities with the Gaussian method proposed by Rui-Ping Li et al. [31] regarding the membership degree computing. Therefore, we can state that our proposals have a distinct physical meaning and well-defined mathematical characteristics [14,15,31].
Clustering has been used in various fields, such as image processing, data mining, pattern recognition, and statistical analysis. Generally, clustering algorithms consider all variables equally relevant or not correlated. Nevertheless, the pattern of data samples in the multidimensional space can be geometrically complicated, e.g., clusters may exist in different subsets of features. In this regard, new soft subspace clustering algorithms have been proposed, in which the correlation and relevance of variables are considered to improve their performance. Since regularization-based methods are robust for initializations, the approaches proposed introduce an entropy regularization term for controlling the membership degree of the objects. Such regularizations are popular due to high performance in large-scale data clustering and low computational complexity. These three-step iterative algorithms provide a fuzzy partition, a representative for each cluster, and the relevance weight of the variables or their correlation by minimizing a suitable objective function. Several experiments on synthetic and real datasets, including their application to the segmentation of noisy image textures, demonstrate the usefulness of the proposed clustering methods.
Soft subspace clustering of interval-valued data with regularizations
2021, Knowledge-Based Systems
Citation Excerpt :
□ It is worth noting that the maximum entropy clustering algorithms are similar to the Gaussian method proposed by Rui-Ping Li et al. [25]. Therefore, it not only has a distinct physical meaning but also possesses well-defined mathematical characteristics [15,17,25].
Data analysis plays an indispensable role in understanding different phenomena. One of the vital means of handling these data is to group them into a set of clusters given a measure of similarity. Usually, clustering methods deal with objects described by single-valued variables. Nevertheless, this representation is too restrictive for representing complex data, such as lists, histograms, or even intervals. Furthermore, in some problems, many dimensions are irrelevant and can mask existing clusters. In this regard, new interval-valued data clustering methods with regularizations and adaptive distances are proposed. These approaches consider that the boundaries of the interval-valued variables have the same and different importance for the clustering process. The algorithms optimize an objective function alternating three steps for obtaining the representatives of each group, a fuzzy partition, and the relevance weights of the variables. Experiments on synthetic and real data sets corroborate the robustness and usefulness of the proposed methods.
SMKFC-ER: Semi-supervised multiple kernel fuzzy clustering based on entropy and relative entropy
2021, Information Sciences
The present study aimed to present a new algorithm called Semi-supervised Multiple Kernel Fuzzy Clustering based on Entropy and Relative entropy (SMKFC-ER) by focusing on external knowledge related to the labeled data. In the proposed method, entropy coefficient and relative entropy divergence measure are applied instead of fuzzifier for unsupervised section and the geometric distance measure for semi-supervised section respectively, by emphasizing on combining unsupervised and semi-supervised sections explicitly. The use of relative entropy and entropy in the objective function results in sharing more consistent concepts for the semi-supervised section, controlling the fuzziness of the extracted clusters, and determining the kernel weights regularly for the unsupervised section. Finally, using relative entropy with entropy simultaneously derives a closed-form solution. The performance and supremacy of the proposed method on non-spherical synthetic and real-world datasets are shown by comparing unsupervised and semi-supervised fuzzy clustering methods.

View all citing articles on Scopus

View full text

Gaussian clustering method based on maximum-fuzzy-entropy interpretation

Abstract

Inform. Control

On Measures of Information and their Characterizations

Pattern Recognition with Fuzzy Objective Function Algorithms

Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of image

IEEE Trans. Pattern Anal. Machine Intell.

Information theory and statistical mechanics

Phys. Rev.

Information theory and statistical mechanics

Phys. Rev.