© 2014, IJARCSMS All Rights Reserved 71 | P age ISSN: 2321-7782 (Online) Volume 2, Issue 3, March 2014 International Journal of Advance Research in Computer Science and Management Studies Research Article / Paper / Case Study Available online at: www.ijarcsms.com Differential Evolution algorithm with Support Vector Machine to classify objects efficiently Sharmishta Desai 1 Research Student Computer Engineering Department VIT, Pune University Pune – India Dr. S. T. Patil 2 Professor Computer Engineering Department VIT, Pune University Pune – India Abstract: This paper explains a method of classifying objects efficiently using optimized search space. Automatic detection and classification of objects is required in many applications like automatic selling of goods, vending machines etc. These applications require extracting monetary characteristics from an object image. To detect and classify objects, many classification algorithms are available. Support vector machines are an effective tool used for accurate classification and detection of objects. Instead of searching objects in whole search space differential evolution algorithm is used to reduce the search space .In this paper, advantages of using differential algorithm and support vector machines instead of using other optimization and classification techniques is explained herewith. Keywords: SVM, DE, Counterfeit Objects, Optimization, Classification, Neural Network, Kernel function. I. INTRODUCTION With the advance of digital imaging technologies, colour scanners and laser printers make it increasingly easier to produce counterfeit objects with high resolution. The proliferation of Counterfeit objects in circulation leads to profit loss of traders and business. Therefore, finding an efficient method to detect counterfeit objects is an imperative and demanding task for business transactions in our daily life. Automatic methods for object recognition are required in many applications such as automatic selling-goods and vending machines. Extracting sufficient monetary characteristics from the object image is essential for accuracy and robustness of the automated system. This is a challenging issue to system designer. Classification problems becoming extremely important in decision science. Differential evolution (DE) is a method that optimizes a problem by iteratively trying to improve a candidate solution with regard to a given measure of quality. Such methods are commonly known as metaheuristics as they make few or no assumptions about the problem being optimized and can search very large spaces of candidate solutions. In this paper, Section II covers work done by different authors in literature. Section III explains proposed method with explanation of support vector machine and differential evolution algorithm. Section IV gives advantages of applying SVM and DE to classification and optimization objects. Section V concludes the paper. II. LITERATURE SURVEY The Turkish Lira and Cyprus Pound et al explained the implementation of a single NN to classify objects. In this paper, author has implemented back propagation Neural network using an input layer with 100 neurons, one hidden layer with 30 neurons and one combined output layer with nine neurons. But the neural network requires sufficient amount of training data. As data increases, the number of neurons increases and number of layers also increases with this. It takes more execution time. Hamid Hassanpour , Payam M. Farahabadi et al.[2] explained using Hidden Markov Models for paper currency recognition. Hidden markov model is recursive algorithm so it expensive in terms of memory and time. Fumiaki Takeda, S. Omatu et al. [1] explained A neuro-paper currency recognition method using optimized masks by genetic algorithm. Masks are used for Paper