Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Sequential monte carlo samplers

About

This paper shows how one can use Sequential Monte Carlo methods to perform what is typically done using Markov chain Monte Carlo methods. This leads to a general class of principled integration and genetic type optimization methods based on interacting particle systems.

Pierre Del Moral, Arnaud Doucet• 2002

Related benchmarks

TaskDatasetResultRank
Unconditional modelingFunnel d = 10
Delta log Z0.561
30
Unconditional modeling25GMM d = 2
Delta Log Z0.569
30
Unconditional modelingManywell d = 32
Δ log Z14.99
29
Toy target distribution samplingRings d = 2
Entropy-Reg W2 (eps=0.05)0.18
7
Toy target distribution samplingFunnel d = 10
KS Distance0.035
7
Bayesian Logistic RegressionSonar d=34
Avg. Posterior Log-Likelihood-111
7
Bayesian Logistic RegressionIonosphere (d=61)
Avg Posterior Log-Likelihood-87.82
7
Bayesian Logistic RegressionIonosphere d = 35 (test)
Predictive Likelihood-87.79
7
Sampling toy distributions8-Gaussians (d=2)
2-Wasserstein Distance (Entropic Reg.)0.99
7
Bayesian Logistic RegressionSonar d = 61 (test)
Predictive Likelihood-110.9
7
Showing 10 of 12 rows

Other info

Follow for update