Implementation of Support Vector Machines on FPGA
Support Vector Machine, Sequential Minimal Optimization, FPGA.
The project’s main goal is the parallel FPGA implementation of both the feed-forward phase of a Support Vector Machine (SVM) as well as its trainning phase. The training phase of the SVM will be implemented using Sequential Minimal Optimization (SMO) and the kernel used will be the polynomial. After the parallel implementation in hardware, the SVM will be validated by simulation. Moreover, analysis associated with the temporal performance of the proposed structure, as well as analysis associated with FPGA’s, area will be performed.