AI Paper Atlas

Research map / Detection and segmentation / CNN features

Fast Online Changepoint Detection via Functional Pruning CUSUM Statistics

Gaetano Romano, Idris A. Eckley, Paul Fearnhead, Guillem Rigaill

NeurIPS 2023 · Detection and segmentation › CNN features · approach: Region-based and query-based detectors · 10 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 FOCuS (functional pruning, all change sizes) instead of Numenta HTMR-FOCuS · NAB AWS CPU utilizationprecision: 0.50 -> 0.58Improved
use R-FOCuS instead of Numenta HTMR-FOCuS · NAB AWS CPU utilizationrecall: 0.76 -> 0.82Improved
use R-FOCuS instead of Numenta HTMR-FOCuS · NAB AWS CPU utilizationfalse detections: 13 -> 7Improved
use FOCuS instead of ocd, pre-change mean unknown, large changeFOCuS · 100 streams sparsity 1, magnitude 1average detection delay: 675.58 -> 190.29Improved
use FOCuS0 instead of ocd, pre-change mean known, small changeFOCuS0 · 100 streams sparsity 0.01, magnitude 0.25average detection delay: 318.30 -> 365.03Hurt
use FOCuS0 instead of ocd, pre-change mean known, dense large changeFOCuS0 · 100 streams sparsity 1, magnitude 2average detection delay: 17.85 -> 39.74Hurt
use FOCuS0 instead of ocd, pre-change mean known, sparse large changeFOCuS0 · 100 streams sparsity 0.01, magnitude 2average detection delay: 12.65 -> 7.34Improved
use FOCuS (all mu1) instead of Page-CUSUM on a 10-point gridFOCuS0 · Gaussian mean change simulationdetection delay for off-grid change sizesImproved
use FOCuS instead of MOSUMFOCuS0 · Gaussian mean change simulationdetection delayImproved
prune FOCuS0 to a 10-point grid (FOCuS0-10p) from exactFOCuS0 · Gaussian mean change simulationdetection delayNo clear effect
use FOCuS (unknown mean) instead of FOCuS0 with small training dataFOCuS · Gaussian mean change simulationdetection delayImproved
use FOCuS instead of FOCuS0 with large training dataFOCuS · Gaussian mean change simulationdetection delayMixed

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