USE OF FPGA FOR REAL-TIME K-MEANS CLUSTERİNG ALGORITHM

Authors : Muhammed Yildirim, Ahmet Çinar

Pages : 130-136

View : 16 | Download : 12

Publication Date : 2019-09-28

Article Type : Research

Abstract :Data mining is important methods in the data processing step. Due to the fact that computer technologies are becoming increasingly cheap and their power is increasing day by day, they allow computers to store data in larger quantities [1]. Owing to the improving of technology, many transactions are recorded in an electronic device and this records can be safely stored. This data can easily be accessed when requested. By means of developing technologies, it is ensured that these processes are getting more day by day at a lower cost. Therefore, it is of great importance to be able to process these data of high size. Clustering algorithms that aggregate the data in the database under groups or clusters to bring together objects with similar properties have a great deal of data mining proposition. In this paper, it is aimed to collect 2 clusters based on the similarities of 60 data obtained from 2 different wheat varieties using k-means clustering algorithm based on Fpga architecture. Since the FPGA architecture has the ability to perform parallel processing, it will shorten the processing time and so efficiency will increase. Also, the ability to use FPGA’s over and over again provides an extra advantage. The proposed system is designed using the verilog hardware identification language on the DE2_115 Fpga board.
Keywords : Data Mining, Fpga, K-means, Verilog