115. What is machine learning?

Many people may have already asked this question: What is machine learning?

Arthur Samuel (1959) said that Machine Learning is a field of study which gives computers the ability to learn without being explicitly programmed.

The above answer is not satisfactory. For example, people may further raise these questions such as:

  1. What is the generalized meaning behind the word “learning”?
  2. What is the generalized meaning behind the word “programming” or “being programmed”?
  3. When a person receives a series of instructions such as “take your dinner”, “go to walk outside for 10 minutes”, “do study for 1 hour”, “prepare for sleeping”, etc., is this person actually being programmed or not in a broad sense?

Hence, people are still expecting better definitions of machine learning. Then, how should we define machine learning in a better way?

Here is my attempt of defining machine learning from a different viewpoint:

Machine Learning is a discipline which studies the principles underlying data analysis for the purpose of enabling the mental activities of cognition, memorization and recognition of patterns (including models) such as relational patterns, spatial patterns, time-varying patterns and behavioral patterns, etc.

If you have definitions of machine learning, which are better, please share your versions or thoughts.

Ming Xie

Nanyang Technological University