1. A NOVEL APPROACH FOR LEARNING RATE IN SELF ORGINIZING MAP (SOM)
- Author
-
Emin Germen, Anadolu Üniversitesi, Mühendislik Fakültesi, Elektrik ve Elektronik Mühendisliği Bölümü, and Germen, Emin
- Subjects
Taşınım ,Akustik ,Computer science ,Ekoloji ,Mühendislik ,Astronomi ve Astrofizik ,Kimya ,Engineering ,Mikroskopi ,Self Organizing Map,Kohonen,Learning Rate ,lcsh:Technology (General) ,Tıbbi ,İnorganik ve Nükleer ,Kohonen ,Partiküller ve Alanlar ,TRACE (psycholinguistics) ,Matematik ,Mekanik ,Kristalografi ,Spektroskopi ,Zooloji ,Entomoloji ,Fizikokimya ,Nükleer ,Organik ,General Medicine ,Toksikoloji ,Genetik ve Kalıtım ,Termodinamik ,Mantar Bilimi ,Su Kaynakları ,Analitik ,Self Organizing Map ,Katı Hal ,Biyoloji ,Balıkçılık ,Matching (graph theory) ,Deniz ve Tatlı Su Biyolojisi ,Structure (category theory) ,Fizik ,Network topology ,Learning Rate ,Synthetic data ,Kuş Bilimi ,Uygulamalı ,İstatistik ve Olasılık ,Evrim Biyolojisi ,Ortak Disiplinler ,Training period ,Limnoloji ,Mineraloji ,Farmakoloji ve Eczacılık ,business.industry ,Kohonen self organizing map ,Pattern recognition ,Rate parameter ,Akışkanlar ve Plazma ,Optik ,Çevre Bilimleri ,lcsh:TA1-2040 ,Biyoliji Çeşitliliğinin Korunuması ,Disiplinler Arası Uygulamalar ,lcsh:T1-995 ,Artificial intelligence ,Karşılaştırmalı Bioloji ,lcsh:Engineering (General). Civil engineering (General) ,business ,Atomik ve Moleküler Kimya - Abstract
The performance of resultant topological structure of Kohonen Self Organizing Map SOM is highly dependent of the learning rate and neighborhood parameters. In literature there are plenty many different types of approaches to and proposals for those parameters. It has been investigated that in general the learning rate and neighborhood parameters are data independent and predefined before the training period. Here in this paper a novel approach has been proposed to change the learning rate parameter according to the interaction of neurons with data. During training, the worst matching neuron also tracked and used to trace the formation of topological structure of SOM. A slight modification on conventional learning rate with proposed method has a considerable influence on resultant topologies in a positive way. The effects of this approach have been tested with the real world problem and different synthetic data.
- Published
- 2018
- Full Text
- View/download PDF