In this paper, we propose a new type of supervised multi-layered self-organizing map and examine to what extent information content in multi-layered ...
Jul 1, 2019 · The present paper aims to show the necessity of information augmentation to cope with the natural decrease in information content in multi-layered neural ...
The self-organizing map (SOM) is one of the popular clustering and data visualization algorithms and has evolved as a useful tool in pattern recognition, ...
In this method, the self-organizing map (SOM) is used to produce the knowledge (SOM knowledge) on input patterns. SOM knowledge is sometimes redundant and not ...
Abstract—In this paper, we introduce a new type of information-theoretic method called “information-theoretic active. SOM”, based on the self-organizing ...
In this paper, we propose a new information-theoretic method called “enhancement learning” to interpret the configuration of competitive networks.
21 Citations · Information-theoretic multi-layered supervised self-organizing maps for improved prediction performance and explicit internal representation.
People also ask
Is self organizing maps supervised or unsupervised?
What is advantage of self organizing maps when compared to neural networks?
Sep 23, 2015 · This paper proposes a new information-theoretic method based on the information enhancement method to extract important input variables.
We introduce the Supervised Self-organizing Maps (SuSi) framework, which can perform unsupervised, supervised and semi-supervised classification as well as ...
Feb 15, 2022 · A collection of MATLAB modules for self organizing maps and counterpropagation artificial neural networks.