“Decoding candidemia in critically ill patients: unsupervised clustering identifies three unique phenotypes”: algorithmic tautology and feature selection bias
Zejun Yang, Yaotian Zhang · 2026
Read “Decoding candidemia in critically ill patients: unsupervised clustering identifies three unique phenotypes”: algorithmic tautology and feature selection bias by Zejun Yang, Yaotian Zhang (2026) 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-13523429.txt.
Description
Zejun Yang and Yaotian Zhang write a 2026 letter that questions a recent study of a blood infection in very ill patients. The study they discuss used machine learning to sort patients with candidemia into three groups with quite different death rates. Yang and Zhang praise the aim but argue the design has deep flaws that limit its use in real care.
Their first charge is that the sorting restates what the input scores already say, so the groups add little new insight. Their second charge is bias in how traits were picked for the model. The letter urges more care before such groups guide trials or treatment.
Catalog record
| CREATOR | Zejun Yang, Yaotian Zhang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 7 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |