AI Paper Atlas

Research map / Optimization / Optimizers

The Devil is in the Detail: A Framework for Macroscopic Prediction via Microscopic Models

Yingxiang Yang, Negar Kiyavash, Le Song, Niao He

NeurIPS 2020 · Optimization › Optimizers · approach: Randomized gradient estimation · 4 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 CSO with GDRS instead of Macro-SLmicro-model with SIR · COVID-19 county datatest loss: 0.086 -> 0.041Improved
Use CSO with GDRS instead of Macro-SLmicro-model with Neural ODE · COVID-19 county datatest loss: 0.037 -> 0.013Improved
Use CSO with GDRS instead of Micro-SLmicro-model with Neural ODE · COVID-19 county datatest loss: 0.049 -> 0.013Improved
Use GDRS instead of biased SGD to solve CSOSIR micro-model · COVID-19 county datatest loss: 0.044 -> 0.041Improved
Use GDRS instead of biased SGD to solve CSONeural ODE micro-model · COVID-19 county datatest loss: 0.013 -> 0.013No clear effect
Increase the number of random-search particles m (1 to 1000) (ablation)GDRS · toy linear regressionl2 test loss convergence speedImproved
Use GDRS instead of biased SGD and Macro-SLCSO · toy linear regressionl2 test lossImproved
Use Micro-SL instead of GDRS (needs micro labels)toy linear regression · toy linear regressionl2 test lossMixed
Use GDRS instead of competing algorithms under noisy training data and small aggregation ratioCSO · Medicare SynPUFl2 and KL test lossImproved

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