首页|Research Findings from University of Wisconsin Madison Update Understanding of M achine Learning (Symbolic Regression on FPGAs for Fast Machine Learning Inferenc e)
Research Findings from University of Wisconsin Madison Update Understanding of M achine Learning (Symbolic Regression on FPGAs for Fast Machine Learning Inferenc e)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Researchers detail new data in artific ial intelligence. According to news reporting out of the University of Wisconsin Madison by NewsRx editors, research stated, "The high-energy physics community is investigating the potential of deploying machine-learning-based solutions on Field-Programmable Gate Arrays (FPGAs) to enhance physics sensitivity while stil l meeting data processing time constraints." The news correspondents obtained a quote from the research from University of Wi sconsin Madison: "In this contribution, we introduce a novel end-to-end procedur e that utilizes a machine learning technique called symbolic regression (SR). It searches the equation space to discover algebraic relations approximating a dat aset. We use PySR (a software to uncover these expressions based on an evolution ary algorithm) and extend the functionality of hls4ml (a package for machine lea rning inference in FPGAs) to support PySR-generated expressions for resource-con strained production environments. Deep learning models often optimize the top me tric by pinning the network size because the vast hyperparameter space prevents an extensive search for neural architecture. Conversely, SR selects a set of mod els on the Pareto front, which allows for optimizing the performance-resource tr ade-off directly. By embedding symbolic forms, our implementation can dramatical ly reduce the computational resources needed to perform critical tasks."
University of Wisconsin MadisonCyborgsEmerging TechnologiesMachine Learning
NETL
NSTL
