Incremental grid growing: encoding high-dimensional structure into a two-dimensional feature map

Blackmore, Justine; Miikkulainen, Risto

doi:10.1109/icnn.1993.298599

Cited by 85 publications

(48 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are many approaches which apply these algorithms in classic neural networks (Islam et al, 2009), (Bortman and Aladjem, 2009), (Han and Qiao, 2013), (Yang and Chen, 2012). Also, there are many variations of SOM that allow a more flexible structure of the output map which can be divided into two categories: In the first type, we include growing grid (GG) , incremental GG (Blackmore and Miikkulainen, 1993), growing SOM (GSOM) (Alahakoon et al, 2000) all coming with different variants. GG is the only variant which allows growing a new node from the interior of the grid (but this is a whole row or column of nodes).…”

Section: Flexible Structure In Neuralmentioning

confidence: 99%

AMSOM: Adaptive Moving Self-organizing Map for Clustering and Visualization

Spanakis¹,

Weiß

2016

Proceedings of the 8th International Conference on Agents and Artificial Intelligence

View full text Add to dashboard Cite

Abstract:Self-Organizing Map (SOM) is a neural network model which is used to obtain a topology-preserving mapping from the (usually high dimensional) input/feature space to an output/map space of fewer dimensions (usually two or three in order to facilitate visualization). Neurons in the output space are connected with each other but this structure remains fixed throughout training and learning is achieved through the updating of neuron reference vectors in feature space. Despite the fact that growing variants of SOM overcome the fixed structure limitation they increase computational cost and also do not allow the removal of a neuron after its introduction. In this paper, a variant of SOM is proposed called AMSOM (Adaptive Moving Self-Organizing Map) that on the one hand creates a more flexible structure where neuron positions are dynamically altered during training and on the other hand tackles the drawback of having a predefined grid by allowing neuron addition and/or removal during training. Experiments using multiple literature datasets show that the proposed method improves training performance of SOM, leads to a better visualization of the input dataset and provides a framework for determining the optimal number and structure of neurons.

show abstract

Section: Flexible Structure In Neuralmentioning

confidence: 99%

AMSOM: Adaptive Moving Self-organizing Map for Clustering and Visualization

Spanakis¹,

Weiß

2016

Proceedings of the 8th International Conference on Agents and Artificial Intelligence

View full text Add to dashboard Cite

show abstract

“…The initial choice of the number of classes is arbitrary, and there does not exist any method to choose the size of the grid in a perfect way, even if some authors suggest using growing or shrinking approaches (Blackmore and Mikkulainen, 1993, Fritzke, 1994, Koikkalainen and Oja, 1990. To exploit the stochastic features of the SOM algorithm, and to obtain a good clustering and a good organization, it is proven to be more efficient to deal with "large" maps.…”

Section: Example I: the Country Database Pop_96mentioning

confidence: 99%

SOM-based algorithms for qualitative variables

Cottrell¹,

Ibbou²,

Letrémy³

2004

Neural Networks

View full text Add to dashboard Cite

It is well known that the SOM algorithm achieves a clustering of data which can be interpreted as an extension of Principal Component Analysis, because of its topologypreserving property. But the SOM algorithm can only process real-valued data. In previous papers, we have proposed several methods based on the SOM algorithm to analyze categorical data, which is the case in survey data. In this paper, we present these methods in a unified manner. The first one (Kohonen Multiple Correspondence Analysis, KMCA) deals only with the modalities, while the two others (Kohonen Multiple Correspondence Analysis with individuals, KMCA_ind, Kohonen algorithm on DISJonctive table, KDISJ) can take into account the individuals, and the modalities simultaneously.

show abstract

“…During the on-line stage, the quantization error measured on each input vector controls the growing process. Related works on growing Kohonen's map have been proposed in [11,8]. The allocation process of new recognition codes (codebooks) is controlled by a threshold value that is empirically found.…”

Section: Self Organizing Mapsmentioning

confidence: 99%