Supporting Data and Systems Analysis with Approximate Feature Selection
Prof. Qiang Shen
Prof. Qiang Shen
Abstract:
Feature selection (FS) addresses the challenge of identifying system descriptors that best predict a given outcome. Unlike other dimensionality reduction methods, FS preserves the original meaning of features. This approach has proven successful in various real-world applications, dealing with datasets containing an extensive number of imprecisely described features, which might be impractical to model and process (e.g., large-scale image analysis, text processing, and web content classification) while needing to retain feature semantics.
This presentation will delve into the development and application of approximate FS mechanisms based on rough and fuzzy-rough theories. These techniques offer an effective means to perform FS without relying on user-supplied information. Specifically, fuzzy-rough feature selection (FRFS) accommodates discrete and real-valued noisy data, making it suitable for both regression and classification. The only additional information required is the fuzzy partition for each feature, which can be automatically derived from available domain data. FRFS has demonstrated its potency as a technique for semantics-preserving data dimensionality reduction.
Starting with an overview of the general background of FS, this talk will first cover the rough-set-based approach before delving into FRFS and its practical applications. The presentation will conclude with an outline of opportunities for further development.
Feature selection (FS) addresses the challenge of identifying system descriptors that best predict a given outcome. Unlike other dimensionality reduction methods, FS preserves the original meaning of features. This approach has proven successful in various real-world applications, dealing with datasets containing an extensive number of imprecisely described features, which might be impractical to model and process (e.g., large-scale image analysis, text processing, and web content classification) while needing to retain feature semantics.
This presentation will delve into the development and application of approximate FS mechanisms based on rough and fuzzy-rough theories. These techniques offer an effective means to perform FS without relying on user-supplied information. Specifically, fuzzy-rough feature selection (FRFS) accommodates discrete and real-valued noisy data, making it suitable for both regression and classification. The only additional information required is the fuzzy partition for each feature, which can be automatically derived from available domain data. FRFS has demonstrated its potency as a technique for semantics-preserving data dimensionality reduction.
Starting with an overview of the general background of FS, this talk will first cover the rough-set-based approach before delving into FRFS and its practical applications. The presentation will conclude with an outline of opportunities for further development.
Qiang Shen received a PhD in Computing and Electrical Engineering (1990) and a DSc in Computational Intelligence (2013). He holds the Established Chair of Computer Science and is Pro Vice-Chancellor: Faculty of Business and Physical Sciences at Aberystwyth University. He is a Fellow of the Royal Academy of Engineering and a Fellow and Council Member of the Learned Society of Wales. He had the honour of being a London 2012 Olympic Torch Relay torchbearer, selected to carry the Olympic torch as part of the centenary celebration of Alan Turing. Also, he has served as a panel member for the past two UK Research Excellence Framework (REF) exercises: 2014 and 2021, on Computer Science and Informatics. He is the recipient of the 2024 IEEE Fuzzy Systems Pioneer Award. Professor Shen has authored 3 research monographs and over 470 peer-reviewed papers. His publications include many outstanding journal articles and best conference papers, which are directly related to the subject matter discussed in this presentation.
|