AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Iterated Block Particle Filter for High-dimensional Parameter Learning: Beating the Curse of Dimensionality

Ning Ning, Edward L. Ionides

JMLR 2023 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Structured Gaussian process models · 11 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use IBPF instead of IF2 for parameter learningIF2 · Measles case 4, 2 cities (BPF metric)log-likelihood: -4373 -> -4358Improved
use IBPF instead of IEnKF for parameter learningIEnKF · Measles case 4, 2 cities (BPF metric)log-likelihood: -70118 -> -4358Improved
use IBPF instead of IF2 under the PF metricIF2 · Measles case 4, 2 cities (PF metric)log-likelihood: -4358 -> -4380No clear effect
use IBPF instead of IF2 under the EnKF metricIF2 · Measles case 4, 2 cities (EnKF metric)log-likelihood: -4970 -> -4929Improved
increase block size from 1-4 cities to 6 cities (ablation)IBPF · Block size study, 12 citieslog-likelihood: {-75793, -96487, -88973, -84087, -76356} for 20-100 iterations, much smaller than block sizes 1-4Hurt
decrease block size from 4 to 1 city per block (ablation)IBPF · Block size study, 12 citieslog-likelihoodImproved
increase number of IBPF iterations from 20 to 180 (ablation)IBPF · Block size study, 12 citieslog-likelihoodImproved
add more cities (dimension) to learn for IF2IF2 · Measles case 2log-likelihood per city per stepHurt
make the parameter learning problem highly nonlinear (cases 3 and 4)IEnKF · Measles cases 3 and 4log-likelihoodHurt
add the highly nonlinear parameter alpha and other location parameters (cases 3 and 4)IBPF · Measles cases 3 and 4log-likelihood vs true parameterNo clear effect
scale number of cities from 2 to 10 with IBPFIBPF · Measles case 4 replicateslog-likelihood per city per step: -5.947 -> -5.956No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy