An eigenvalue-based small-sample approximation to MCMC reduces required paths from up to 1,000,000 to as few as 10 while producing comparable steady-state distributions by Wasserstein distance and lower variance.
Local Graph Partitioning using PageRank Vectors
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
econ.EM 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Fast Monte-Carlo
An eigenvalue-based small-sample approximation to MCMC reduces required paths from up to 1,000,000 to as few as 10 while producing comparable steady-state distributions by Wasserstein distance and lower variance.