On the Perceptron's Compression
arXiv:1806.05403v11 citations
Originality Synthesis-oriented
AI Analysis
This work addresses incremental improvements in training algorithms for machine learning practitioners.
The paper investigates modifications to the perceptron algorithm to improve margin guarantees and applies these insights to neural network training, supported by experimental data.
We study and provide exposition to several phenomena that are related to the perceptron's compression. One theme concerns modifications of the perceptron algorithm that yield better guarantees on the margin of the hyperplane it outputs. These modifications can be useful in training neural networks as well, and we demonstrate them with some experimental data. In a second theme, we deduce conclusions from the perceptron's compression in various contexts.