Self Organizing Map (SOM) is a generic method that has been applied in many applications. I spent almost 40-60 percent of my time over last three years on playing with this computational machine. Today while I was writing my PhD thesis I was trying to find a unique story line for SOM that fits well with the concept of “Pre-specific Modeling”, which is the title of my thesis. I have several technical writings to explain it in detail, but it is not convincing. As a result, I decided that may be it is better not to find a fixed way to explain SOM and maybe it should remain “pre-specific” and in this case “observer dependent”. Here, I collected a set of key words about SOM.

In principle, I can write about all of these indexes or refer to those people and papers who are focused on each of them, but now I prefer leave them as a list. Just to mention that now I am more interested in the third part of indexes which goes into analogies between SOM and other concepts.

**Functions**
*Vector Quantization*
*Dimensionality Reduction*
*Manifold Learning*
*Topology Preservation*
*Representation Learning*
*Computing with Contextual Numbers*
*Multi-dimensional Sorting*
*Nonlinear Eigen form*
*Unsupervised Learning*
*Space Transformation*
*Visualization of High Dimensional Data*
*Data Reduction and Abstraction*
*Clustering*
*Prototype Generation and Emergence*
*Classification*
*Prediction and Function Approximation and Time Series*
*Structure Learning*
*Multi-criteria Optimization*
*Reinforcement Learning*
*Nonparametric Joint Probability Distribution and Resampling*
*Topological Data Analysis *

**Types**
*Fixed Topology*
*One-dimensional SOM*
*Two-dimensional SOM*
*Spherical SOM*
*Parametric SOM*
*Growing SOM*
*Hierarchical SOM*
*Neural Gas*
*Generative Topographic Map*
*Vector Quantization*
*Mixture of Gaussians*
*SOM as a Two layer Neural Network*
*SOM as Radial Basis Function*
*Online Learning*
*Batch Learning*
*Recurrent SOM*
*Recursive SOM*
*Relational SOM*
*Median SOM*
*Semantic SOM*
*WEBSOM*
*Vectorial representation or Non-vectorial SOM*
*Similarity Measures*
*U-Matrix*
*P-Matrix *

**Analogy**
*Associative Memory*
*Brain Cortex Analogy*
*Wave Particle Dualism*
*Mind-Matter Dualism*
*Cartesian Dualism*
*Totalitarian Geometry, Democratic Polynomials and Social Computing of SOMs *
*Geometry: Global Structures, No Local Adaptation, Polynomials: Global structure, Local Adaptation, SOM: Local Structure Local Adaptation*
*Emergent SOM: Engendering Probabilistic Concepts by Increasing the Population of Samples vs. Data Reduction Toward Ideal Representatives*

### Like this:

Like Loading...

*Related*