science1 distinct publisher
Odor mixtures get a benchmark: 0.08 median error, 0.57 correlation, 96 pairs of evidence
A Monell-co-authored DREAM challenge turned odor-mixture similarity into a public machine learning task. The winning ensemble's error and its correlation tell different stories about how close this is.
Publishers:phys.org
Reality
- Evidence58
- Adoption24
- Hype gap+28
- Incentives62
- Confidence46