Combination of Growing and Pruning Algorithms for Multilayer Perceptrons for Speech/Music/Noise Classification in Digital Hearing Aids

Álvarez L, Alexandre E, Llerena C, Gil-Pita R, Rosa-Zurera M. Combination of Growing and Pruning Algorithms for Multilayer Perceptrons for Speech/Music/Noise Classification in Digital Hearing Aids. In: AES Convention 134. Audio Engineering Society; 2013. Paper 8850. Available from: https://aes.org/publications/elibrary-page/?id=16751

@inproceedings{Alvarez2013_16751,
  author = {Álvarez, Lorena and Alexandre, Enrique and Llerena, Cosme and Gil-Pita, Roberto and Rosa-Zurera, Manuel},
  title = {{Combination of Growing and Pruning Algorithms for Multilayer Perceptrons for Speech/Music/Noise Classification in Digital Hearing Aids}},
  booktitle = {AES Convention 134},
  note = {Paper 8850},
  year = {2013},
  month = may,
  publisher = {Audio Engineering Society},
  url = {https://aes.org/publications/elibrary-page/?id=16751}
}

TY  - CPAPER
TI  - Combination of Growing and Pruning Algorithms for Multilayer Perceptrons for Speech/Music/Noise Classification in Digital Hearing Aids
AU  - Álvarez, Lorena
AU  - Alexandre, Enrique
AU  - Llerena, Cosme
AU  - Gil-Pita, Roberto
AU  - Rosa-Zurera, Manuel
T2  - AES Convention 134
M1  - Paper 8850
PY  - 2013
DA  - 2013/05/06
UR  - https://aes.org/publications/elibrary-page/?id=16751
PB  - Audio Engineering Society
LA  - en
AB  - This paper explores the feasibility of combining both growing and pruning algorithms in some way that the global approach results in finding a smaller multilayer perceptron (MLP) in terms of network size, which enhances the speech/music/noise classification performance in digital hearing aids, with the added bonus of demanding a lower number of hidden neurons, and consequently, lower computational cost. With this in mind, the paper will focus on the design of an approach that starts adding neurons to an initial small MLP until the stopping criteria for the growing stage is reached. Then, the MLP size is reduced by successively pruning the least significant hidden neurons while maintaining a continuous decreasing function. The results obtained with the proposed approach will be compared with those obtained when using both growing and pruning algorithms separately.
ER  -