임베디드 시스템의 기계학습을 위한 양자화 학습의 확률 미분 방정식
Stochastic Differential Equation of Quantized Learning for Machine Learning on Embedded System
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
In this paper, we propose the stochastic differential equation(SDE) for the quantized learning equation. We analyze the uniform quantized learning equation based on the central limit theorem and yield an overdamped Langevin dynamics that allows us to control the dynamics of the learning equation. As the proposed analysis demonstrates, the experimental results show that quantized learning performance is better than the conventional learning equations in image classification problems based on the global optimization quantity.
Abstract
In this paper, we propose the stochastic differential equation(SDE) for the quantized learning equation. We analyze the uniform quantized learning equation based on the central limit theorem and yield an overdamped Langevin dynamics that allows us to control the dynamics of the learning equation. As the proposed analysis demonstrates, the experimental results show that quantized learning performance is better than the conventional learning equations in image classification problems based on the global optimization quantity.