Unsupervised learning
Machine learning that finds patterns in data with no human labels, forming its own groupings. Used in bank fraud detection and the Netflix recommendation system, which sorts viewers from what they watch, finish, and skip.
A population health team clusters its heart failure patients on labs, medications and admissions, and the software returns five groups, one of which looks like a distinct high-risk phenotype. The number of groups was a setting someone chose, and a different setting or a different year of data can split or merge them. Before a care pathway is built around any group, check whether it reappears when the analysis is rerun on another year, and have cardiologists judge whether its members share something clinically real.
Terms like this come up in real clinical scenarios across the HelloAI courses: bite-sized modules with verifiable certificates. An account takes one minute, no password needed.
Sign in →