AUTOMATIC ADJUSTMENT OF LEARNING RATES OF THE SELF-ORGANIZING FEATURE MAP

Q: How can I download an article?

To download an article from SID, first log in to the site, search for the article title, and click on the 'Download Article' option.

Q: How can I download an ISI article?

To download an ISI article on SID, enter the keyword or article title in the search bar, view the relevant results, click on the desired article, and select the 'Download Article' option.

Q: How can I access the SID database?

To access the SID database, visit SID.ir, create an account, and log in to access scientific resources.

Q: Is downloading articles from SID free?

Some articles on SID are available for free, while others require payment. Details are specified on the article's page.

SHAH HOSSEINI H.; SAFABAKHSH R.

مرکز اطلاعات علمی Scientific Information Database (SID) - Trusted Source for Research and Academic Resources

Journal Paper

Paper Information

Journal: SCIENTIA IRANICA Year:2001 | Volume:8 | Issue:4 (COMPUTER ENGINEERING) Page(s): 277-286

Download Full-Text

View:

366

Download:

265

Cites:

Information Journal Paper

Title

AUTOMATIC ADJUSTMENT OF LEARNING RATES OF THE SELF-ORGANIZING FEATURE MAP

Author(s)

SHAH HOSSEINI H. | SAFABAKHSH R. | Issue Writer Certificate

Keywords

Not Registered.

Abstract

Time-decreasing learning rate and neighborhood functions of the conventional SOFM (Self- Organizing Feature Map) algorithm are two factors that reduce the capability of this map to adapt weights for different environments. Consequently, parameters for each environment have to be selected empirically, which is a very time-consuming process. In this paper, for dealing with non-stationary input distributions and varied environments, a SOFM algorithm is proposed that automatically adjusts the learning rate of each neuron independently. The learning rate is adjusted by the function of distance between an input vector and the weight vectors. The learning rate modification rule for each output neuron maximizes the correlation between a normalized error of the output neuron and its learning rate parameter. It is also demonstrated that learning rates are able to adjust themselves to vary between zero and one according to their surrounding conditions. With these modifications, the proposed SOFM algorithm virtually has no initial condition requirements crucial to its success when it is used as a Vector Quantize (VQ) network. Moreover, the proposed network has some degree of incremental learning capability and converges to a topographic feature map representing the distribution of input vectors. Experimental results illustrate the superiority of the proposed SOFM algorithm in learning samples of non-stationary distributions. They also indicate that the proposed algorithm speeds up the SOFM convergence and stabilizes with lower distortion values. Moreover, it is shown that with a time-decreasing exponential neighborhood function, the proposed SOFM converges to the topographic map of the input samples.

Cites

A NEW STATISTICAL CHARACTERISTICS BASED METHOD FOR ADAPTIVE LEARNING RATE ADJUSTMENT IN LEARNING AUTOMATA

References

No record.

Cite

APA: Copy

SHAH HOSSEINI, H., & SAFABAKHSH, R.. (2001). AUTOMATIC ADJUSTMENT OF LEARNING RATES OF THE SELF-ORGANIZING FEATURE MAP. SCIENTIA IRANICA, 8(4 (COMPUTER ENGINEERING)), 277-286. SID. https://sid.ir/paper/289684/en

Vancouver: Copy

SHAH HOSSEINI H., SAFABAKHSH R.. AUTOMATIC ADJUSTMENT OF LEARNING RATES OF THE SELF-ORGANIZING FEATURE MAP. SCIENTIA IRANICA[Internet]. 2001;8(4 (COMPUTER ENGINEERING)):277-286. Available from: https://sid.ir/paper/289684/en

IEEE: Copy

H. SHAH HOSSEINI, and R. SAFABAKHSH, “AUTOMATIC ADJUSTMENT OF LEARNING RATES OF THE SELF-ORGANIZING FEATURE MAP,” SCIENTIA IRANICA, vol. 8, no. 4 (COMPUTER ENGINEERING), pp. 277–286, 2001, [Online]. Available: https://sid.ir/paper/289684/en

Related Journal Papers

No record.

Related Seminar Papers

No record.

Related Plans

No record.

Recommended Workshops