Tree-structured Ising models can be learned efficiently

Qinxuan Pan
Cited by: 2|Views4

Abstract:

We provide the first polynomial-sample and polynomial-time algorithm for learning tree-structured Ising models. In particular, we show that $n$-variable tree-structured Ising models can be learned computationally-efficiently to within total variation distance~$\epsilon$ from an optimal $O(n \log n/\epsilon^2)$ samples, where $O(.)$ hide...More

Code:

Data:

Full Text
Bibtex