Uncertainty Quantification and Flagging of Unreliable Predictions in Predicting Mass Spectrometry-Related Properties of Small Molecules Using Machine Learning

Dmitriy D. Matyushin, Ivan A. Burov, Anastasia Yu. Sholokhova · 2024

Read Uncertainty Quantification and Flagging of Unreliable Predictions in Predicting Mass Spectrometry-Related Properties of Small Molecules Using Machine Learning by Dmitriy D. Matyushin, Ivan A. Burov, Anastasia Yu. Sholokhova (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11641629.txt.

Description

Dmitriy Matyushin, Ivan Burov, and Anastasia Sholokhova tackle a trust problem in machine learning for chemistry in a 2024 paper. Models can predict traits of small molecules that help identify them from lab spectra. Yet a bare prediction gives no hint of how wrong it might be.

The team built ways to flag shaky predictions. They used the spread of answers across a group of models, how close a new molecule sits to known ones, and which cluster it falls in. Models trained on those signals sorted good calls from the worst ones with fair success across three lab tasks. Flagging weak calls helps chemists know when to seek more proof before naming a substance.

Catalog record

CREATORDmitriy D. Matyushin, Ivan A. Burov, Anastasia Yu. Sholokhova
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC11641629

Browse the full library · About the Tower of Babel