arXiv · cond-mat/9712177
Learning a spin glass: determining Hamiltonians from metastable states
Abstract
We study the problem of determining the Hamiltonian of a fully connected Ising Spin Glass of $N$ units from a set of measurements, whose sizes needs to be ${\cal O}(N^2)$ bits. The student-teacher scenario, used to study learning in feed-forward neural networks, is here extended to spin systems with arbitrary couplings. The set of measurements consists of data about the local minima of the rugged energy landscape. We compare simulations and analytical approximations for the resulting learning curves obtained by using different algorithms.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Silvia Kuva, Osame Kinouchi, Nestor Caticha. 1997-12-15. Learning a spin glass: determining Hamiltonians from metastable states. https://doi.org/10.1016/s0378-4371(98)00126-5
Cite the original work for its findings. Save a collection to share your selection of sources.