Fast self-organizing feature map algorithm

@article{Su2000FastSF,
  title={Fast self-organizing feature map algorithm},
  author={Mu-Chun Su and Hsiao-Te Chang},
  journal={IEEE transactions on neural networks},
  year={2000},
  volume={11 3},
  pages={
          721-33
        }
}
We present an efficient approach to forming feature maps. The method involves three stages. In the first stage, we use the K-means algorithm to select N2 (i.e., the size of the feature map to be formed) cluster centers from a data set. Then a heuristic assignment strategy is employed to organize the N2 selected data points into an N x N neural array so as to form an initial feature map. If the initial map is not good enough, then it will be fine-tuned by the traditional Kohonen self-organizing… 
View on PubMed
cse.unr.edu
114 Citations
Highly Influential Citations
2
Background Citations
19
Methods Citations
35

114 Citations

Citation Type

Citation Type

A Fast Self-Organizing Map Algorithm by Using Genetic Selection
  • He Ni
  • Computer Science
    2009 Third International Symposium on Intelligent Information Technology Application
  • 2009
TLDR
By using a data pre-processing inspired by the genetic selection, the improved self-organizing map algorithm can converge faster than the conventional self- Organizing map in data clustering.
DYNAMIC SUBSET SELECTION APPLIED TO SELF-ORGANIZING MAPS
TLDR
This work is the application of DSS to the SOM, with modifications to both, and shows that the DSSSOM achieves a performance on par with the SOM with training time reduced by a factor of up to nearly five hundred.
Speeding up the Self-Organizing Feature Map Using Dynamic Subset Selection
TLDR
An active learning algorithm is devised for training Self-Organizing Feature Maps on large data sets that introduces definitions for exemplar difficulty suitable to an unsupervised learning context and therefore appropriate Self-organizing map (SOM) stopping criteria.
P1dsom - a Fast Search Algorithm for High-Dimensional Feature Space Problems
TLDR
A novel hierarchical SOM search algorithm is proposed which significantly reduces the expensive computational cost associated with traditional SOM and maintains all merits and qualities of the traditional SOM.
Fast Self-Organizing Maps Training
TLDR
A fast approach to train self-organizing maps that has performance on par with self- Organizing maps trained with the standard approach, but with a significantly reduced training time.
Robust Self-organizing Maps
TLDR
A variant to the learning algorithm that is robust under the presence of outliers in the data by being resistant to these deviations is proposed, called Robust SOM (RSOM).
An Efficient Initialization Scheme for SOM Algorithm Based on Reference Point and Filters
TLDR
The RPSOM dramatically reduces the time complexity, especially in the computation of large data set, and is found that it is better to construct a good initial map and then to use the unsupervised learning to make small subsequent adjustments.
A self organizing map-harmony search hybrid algorithm for clustering biological data
TLDR
This paper proposes an SOM-Harmony Hybrid algorithm that will compute the optimal dimension of the SOM grid with minimum values of topographic and quantization errors.
Fast Competition Approach using Self Organizing Map for Lip-Reading Applications
TLDR
This paper presents a new strategy capable to accelerate the self organizing map's competition algorithm and concerns "only" with the neurons which are aligned along the low-order principal components of the feature space and neglects the rest of neurons.
Improvement of SOM visual stability by adjusting feature maps and sorting of leaning data
  • M. Shinji, M. Tsutomu
  • Computer Science
    The 6th International Conference on Soft Computing and Intelligent Systems, and The 13th International Symposium on Advanced Intelligence Systems
  • 2012
TLDR
Two proposed methods are visually stable than conventional method in the point of feature map location, and the computational complexity of proposed method is greatly reduced.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 43 REFERENCES
Genetic-algorithms-based approach to self-organizing feature map and its application in cluster analysis
  • M. Su, Hsiao-Te Chang
  • Computer Science
    1998 IEEE International Joint Conference on Neural Networks Proceedings. IEEE World Congress on Computational Intelligence (Cat. No.98CH36227)
  • 1998
TLDR
An efficient method is proposed for measuring the degree of topology preservation using genetic algorithms in two stages to form a topologically ordered feature map and a special method is used to interpret a SOFM formed by the proposed GA-based method.
Improving the effectiveness of self-organizing map networks using a circular Kohonen layer
TLDR
A modified, "circular" weight adjustment procedure is introduced and evaluated that is applicable to a class of problems where the actual coordinates of the output map do not need to correspond to the original input topology.
Engineering applications of the self-organizing map
TLDR
The self-organizing map method, which converts complex, nonlinear statistical relationships between high-dimensional data into simple geometric relationships on a low-dimensional display, can be utilized for many tasks: reduction of the amount of training data, speeding up learning nonlinear interpolation and extrapolation, generalization, and effective compression of information for its transmission.
Growing cell structures--A self-organizing network for unsupervised and supervised learning
A multilayer self-organizing feature map for range image segmentation
Application of neural networks in cluster analysis
  • M. Su, N. DeClaris, Ta-Kang Liu
  • Computer Science
    1997 IEEE International Conference on Systems, Man, and Cybernetics. Computational Cybernetics and Simulation
  • 1997
TLDR
A method for automatically discovering the number of clusters and estimating the locations of the centroids of the resulting clusters is proposed based on the interpretation of a self-organizing feature map (SOFM) formed by the given data set.
Training the self-organizing feature map using hybrids of genetic and Kohonen methods
  • M. McInerney, A. Dhawan
  • Mathematics
    Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94)
  • 1994
TLDR
A hybrid algorithm of genetic methods with Kohonen learning that always results in a topologically correct mapping between input and output spaces is described.
Genetic optimization of self-organizing feature maps
  • S. Harp, T. Samad
  • Computer Science
    IJCNN-91-Seattle International Joint Conference on Neural Networks
  • 1991
TLDR
An application of the genetic algorithm to the design of Kohonen self-organizing feature maps using root mean square error and an information-theoretic map entropy measure is presented.
Derivation of a class of training algorithms
  • S. Luttrell
  • Computer Science
    IEEE Trans. Neural Networks
  • 1990
TLDR
A novel derivation is presented of T. Kohonen's topographic mapping training algorithm, based upon an extension of the Linde-Buzo-Gray algorithm for vector quantizer design, which stabilizes the internal representations chosen by a network against anticipated noise or distortion processes.
Globally-ordered topology-preserving maps achieved with a learning rule performing local weight updates only
TLDR
A new unsupervised competitive learning rule is introduced for topology-preserving map formation and vector quantization, which achieves a globally-ordered map by performing local weight updates only, contrary to Kohonen's self-organizing map algorithm and its many variations.
...
1
2
3
4
5
...