Computing Reviews

Uncertainty management with fuzzy and rough sets : recent advances and applications
Bello R., Falcon R., Verdegay J., Springer International Publishing, New York, NY, 2019. 413 pp. Type: Book
Date Reviewed: 06/03/21

This volume contains papers presented at the Second International Symposium on Fuzzy and Rough Sets (ISFUROS 2017). The symposium drew 55 submissions; 30 were accepted for presentation. Extended versions of 20 accepted papers are included in this book, published in Springer’s “Studies in Fuzziness and Soft Computing” series.

The book starts with a well-written preface, which gives the reader a brief introduction to the field. It explains the similarities and differences between fuzzy and rough sets, considering them from two perspectives: as methodologies under the granular computing umbrella, and as members of the soft computing family. The preface also includes information about the conference and an explanatory guide to the papers. A contributor index and list of acronyms can be found before the main part of the book as well.

Most of the contributors are from Latin America. This is not surprising, as one of the main goals of the conference, as stated on its web page, is to increase awareness of fuzzy and rough set theories among the Latin American research community.

The contributions are grouped into three parts: “Fuzzy Sets: Theory and Applications” (nine papers), “Rough Sets: Theory and Applications” (six papers), and “Hybrid Approaches” (five papers). This way of organizing is preferable to putting the papers in alphabetical order. The average length of a paper is 20 pages.

Three keynote speakers presented at the conference; unfortunately, their papers are not included in the book.

Of the nine papers in the first part, two are about fuzzy cognitive maps (FCMs). One of them proposes a combination of FCMs with the computing-with-words (CWW) paradigm, and the other discusses an application of FCMs in software usability evaluation. Two other papers report on applications of fuzzy simulation in production planning and in preventive health support systems, respectively. Other papers cover a hybrid fuzzy clustering algorithm and its application in fault diagnosis; reformulation and a fuzzy logic-based approach to solving “a route planning problem with applications in tourism”; fuzzy characterization of relations between optimal bucket order problem (OBOP) instances and the performance of several algorithms for this problem; two fuzzy linear programming methods (full and mixed integer) for the dynamical and continuous berth allocation problem with two quays and imprecise vessel arrival times; and a new variant of the reference ideal method.

In the part about rough sets, one paper presents “matroidal structures obtained from different partitions and coverings of a specific set,” as well as their generalizations to covering-based rough sets. Other papers are more application oriented. One of them proposes “new methods for solving imbalanced classification problems.” The methods combine rough set theory and the nearest prototype approach. Probabilistic rough sets are used in the early detection of possible undergraduate dropouts and for facial similarity analysis. Other papers report on applications of rough set theory for identifying overlapping communities in social networks and for in-database rule learning in non-stationary environments.

The hybrid approaches are quite diverse. One paper in this part proposes an improvement to rough cognitive ensembles (RCEs) with the help of ordered weighted averaging (OWA) operators. An RCE is “a multiclassifier system composed of a set of rough cognitive networks (RCNs), [where] each [RCN operates] at a different granularity degree.” The improvement concerns “a new activation mechanism for RCE”: a fuzzy strategy, used to activate it before performing the inference process. The strategy requires information aggregation, and this is where the OWA operators are used. Another paper proposes UMDA+RST+FUZZY, a method that facilitates improving the performance of k-nearest neighbors (k-NN) and multilayer perceptron (MLP) algorithms. Other papers describe: an adaptive neuro-fuzzy inference system for scheduling problems in queueing systems with uncertain data; genetic fuzzy systems that help to improve automation in maritime risk assessment; and generalized fuzzy Petri nets with interval triangular norms, which leads to a more realistic model.

As any typical proceedings volume, most won’t read it chapter by chapter. Rather, readers should choose the chapters that interest them--the preface can help to identify those quickly.

Reviewer:  Temur Kutsia Review #: CR147279 (2109-0225)

Reproduction in whole or in part without permission is prohibited.   Copyright 2021™
Terms of Use
| Privacy Policy