STUDY COMPARISON OF SVM-, K-NN- AND BACKPROPAGATION-BASED CLASSIFIER FOR IMAGE RETRIEVAL
Keywords:
Backpropagation, Classification, Image Retrieval, K-NN, SVM
Abstract
Classification is a method for compiling data systematically according to the rules that have been set previously. In recent years classification method has been proven to help many people’s work, such as image classification, medical biology, traffic light, text classification etc. There are many methods to solve classification problem. This variation method makes the researchers find it difficult to determine which method is best for a problem. This framework is aimed to compare the ability of classification methods, such as Support Vector Machine (SVM), K-Nearest Neighbor (K-NN), and Backpropagation, especially in study cases of image retrieval with five category of image dataset. The result shows that K-NN has the best average result in accuracy with 82%. It is also the fastest in average computation time with 17,99 second during retrieve session for all categories class. The Backpropagation, however, is the slowest among three of them. In average it needed 883 second for training session and 41,7 second for retrieve session.
Published
2015-03-26
How to Cite
Athoillah, M., Irawan, M. I., & Imah, E. M. (2015). STUDY COMPARISON OF SVM-, K-NN- AND BACKPROPAGATION-BASED CLASSIFIER FOR IMAGE RETRIEVAL. Jurnal Ilmu Komputer Dan Informasi, 8(1), 11-18. https://doi.org/10.21609/jiki.v8i1.279
Section
Articles
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
