A note on the capacity of the binary perceptron
Abstract
Determining the capacity $\alpha_c$ of the Binary Perceptron is a long-standing problem. Krauth and Mezard (1989) conjectured an explicit value of $\alpha_c$, approximately equal to .833, and a rigorous lower bound matching this prediction was recently established by Ding and Sun (2019). Regarding the upper bound, Kim and Roche (1998) and Talagrand (1999) independently showed that $\alpha_c$ < .996, while Krauth and Mezard outlined an argument which can be used to show that $\alpha_c$ < .847. The purpose of this expository note is to record a complete proof of the bound $\alpha_c$ < .847. The proof is a conditional first moment method combined with known results on the spherical perceptron
- Publication:
-
arXiv e-prints
- Pub Date:
- January 2024
- DOI:
- arXiv:
- arXiv:2401.15092
- Bibcode:
- 2024arXiv240115092A
- Keywords:
-
- Mathematics - Probability;
- Computer Science - Discrete Mathematics;
- Computer Science - Machine Learning;
- Statistics - Machine Learning