Adaptive filtering with the self-organizing map: A performance comparison

• Published in: Neural networks Vol. 19; no. 6; pp. 785 - 798
• Main Authors: Barreto, Guilherme A., Souza, Luís Gustavo M.
• Format: Journal Article Conference Proceeding
• Language: English
• Published: Oxford Elsevier Ltd 01.07.2006; Elsevier Science
• Subjects: Algorithms; Applied sciences; Artificial Intelligence; Computer science; control theory; systems; Computer Simulation; Connectionism. Neural networks; Detection, estimation, filtering, equalization, prediction; Exact sciences and technology; Humans; Information, signal and communications theory; Learning - physiology; Local linear mappings; Miscellaneous; Neural Networks (Computer); Nonlinear Dynamics; Nonlinear equalization; RBF models; Sampling, quantization; Self-organizing map; Signal and communications theory; Signal processing; Signal, noise; Telecommunications and information theory; Vector-quantized temporal associative memory; Vector-quantized temporal associative memory; Local linear mappings; RBF models; Nonlinear equalization; Self-organizing map; Performance evaluation; Input output; Non linear filter; Adaptive filtering; Mapping; Modeling; Learning; Associative memory; Self-organising feature maps; Least squares method; Vector quantization; Adaptive filter; Neural network; Equalization; Non linear filtering; Radial basis function; Linear filter; Transverse filter; Signal processing; Least mean squares methods; Multilayer network
• ISSN: 0893-6080; 1879-2782
• DOI: 10.1016/j.neunet.2006.05.005

In this paper we provide an in-depth evaluation of the SOM as a feasible tool for nonlinear adaptive filtering. A comprehensive survey of existing SOM-based and related architectures for learning input–output mappings is carried out and the application of these architectures to nonlinear adaptive filtering is formulated. Then, we introduce two simple procedures for building RBF-based nonlinear filters using the Vector-Quantized Temporal Associative Memory (VQTAM), a recently proposed method for learning dynamical input–output mappings using the SOM. The aforementioned SOM-based adaptive filters are compared with standard FIR/LMS and FIR/LMS–Newton linear transversal filters, as well as with powerful MLP-based filters in nonlinear channel equalization and inverse modeling tasks. The obtained results in both tasks indicate that SOM-based filters can consistently outperform powerful MLP-based ones.