Design and modeling of parallel support vector machine architecture in printed electronics
Loading...
Date
2025-07-14
Authors
Σερταρίδης, Ηλίας
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Printed Electronics is an emerging technology aimed at enabling low-cost computing applications that are not typically addressed by conventional silicon-based systems. Many of these applications rely on Machine Learning algorithms to perform data classification tasks. Support Vector Machines (SVMs) are one such algorithm, which classify data by separating it in space using a linear decision function. This work introduces a structured framework that trains SVM models and applies optimization techniques to minimize area and power consumption while preserving high classification accuracy. The framework is tested across 10 different datasets, with performance evaluated against relevant state-of-the-art solutions.
Description
Keywords
Printed electronics, Machine learning, Support vector machines, Hardware design, Approximate computing, VLSI