Title :
Effective partitioning of input domains for ALM algorithm
Author :
Afrakoti, I.E.P. ; Ghaffari, Aboozar ; Shouraki, Saeed Bagheri
Author_Institution :
Dept. of Electr. Eng., Sharif Univ. of Technol., Tehran, Iran
Abstract :
This paper presents a new and simple algorithm for partitioning the input domain for implementation of Active Learning Method (ALM) algorithm. ALM is a pattern-based algorithm for soft computing which uses the Ink Drop Spread (IDS) algorithm as its main engine for feature extraction. In this paper a simple algorithm is introduced with a few computation cost. In order to evaluate the performance of the proposed algorithm, it is applied to two applications, system modeling and pattern recognition. Simulation results show the effectiveness of our algorithm in specifying the appropriate points for dividing the inputs domains.
Keywords :
inference mechanisms; learning systems; pattern recognition; ALM algorithm; IDS algorithm; active learning method; feature extraction; ink drop spread algorithm; input domain partitioning; pattern recognition; pattern-based algorithm; soft computing; system modeling; Classification algorithms; Computational modeling; Equations; Feature extraction; Ink; Mathematical model; Partitioning algorithms; Active learning method (ALM); fuzzy inference algorithm; ink drop spread (IDS);
Conference_Titel :
Pattern Recognition and Image Analysis (PRIA), 2013 First Iranian Conference on
Conference_Location :
Birjand
Print_ISBN :
978-1-4673-6204-7
DOI :
10.1109/PRIA.2013.6528437
