COMPARE THE PERFORMANCE OF BACKPROPAGATION AND FORWARD-FORWARD ALGORITHM IN IMAGE RECOGNITION PROBLEM
Keywords:
Backpropagation; Neural Network; Forward-Forward; Forward; Backward.
Abstract
In this article, we will compare the performance of two algorithms,
Backpropagation and Forward-Forward, in image recognition problems using
MNIST data. We will evaluate the accuracy and training speed of both
algorithms. The results show that Backpropagation achieves higher accuracy
when Forward-Forward can have faster training speed in some cases.
điểm /
đánh giá
Published
2025-05-21
Issue
Section
RESEARCH AND DEVELOPMENT