Machine Learning
Also called: ML, statistical learning
A method of building software where the system derives its behavior from patterns in training data rather than from explicitly written rules.
Machine learning builds a model by exposing an algorithm to examples and letting it adjust internal parameters until its outputs match the labels in the training data. The result is a function that maps inputs to outputs without anyone having written the mapping.
In medicine this matters because many clinically useful patterns cannot be written as rules. No one can specify in code what a malignant nodule looks like across every scanner, protocol, and patient. A model trained on enough labeled examples can approximate it.
Where This Gets Misread
The behavior is inherited from the training data, including its errors, its demographic composition, and the scanner or site it came from. A model is not neutral machinery applied to your patients ... it is a compressed summary of somebody else's patients, and how well those two populations match is the whole question.