ML as a subfield of AI and generic term for the "artificial" generation of knowledge, uses computer-based techniques to enable systems to learn from data or experience. Such a system can generalise the acquired knowledge after the end of the learning phase by recognising patterns and regularities from the learning data and transferring them to unknown data (see transfer learning).
See also:ETSI GR ENI 004 set of processes that enables computers to understand data and enhance its knowledge; said knowledge is used to learn new information without being explicitly programmed.
NOTE 1: ISO/IEC 2382-28 defines machine learning as "a process by which a functional unit improves its performance by acquiring new knowledge or skills, or by reorganizing existing knowledge or skills".
NOTE 2: Mitchell's book (Machine Learning) defines this as: "A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E".
NOTE 3: Machine Learning is a subsidiary ongoing application of AI based around the idea that it should give machines access to data and let them learn for themselves.
ISO/IEC DIS 22989 ML: process of optimizing model parameters (3.1.28) through computational techniques, such that the model's behaviour reflects the data or experience
ISO/IEC TR 29119-11:2020 process using computational techniques to enable systems to learn from data or experience
ISTQB - CTAI Syllabus The process using computational techniques to enable systems to learn from data or experience (ISO/IEC TR 29119-11)