Machine learning

noun (uncountable)

/məˈʃiːn ˈlɜːnɪŋ/

A subfield of artificial intelligence in which computer systems improve their performance on a task automatically through experience — by identifying patterns in data — rather than through explicit rule-based programming. The three main paradigms are supervised learning (labelled data), unsupervised learning (unlabelled data), and reinforcement learning (reward-based feedback). Machine learning underpins spam detection, credit scoring, medical diagnosis, weather forecasting, and language translation. In UPSC context, ML is examined in GS3 as a driver of the IndiaAI Mission, judicial AI tools (SUPACE — Supreme Court Portal for Assistance in Courts Efficiency), crop failure prediction, and targeted welfare beneficiary identification under schemes like PM-KISAN.

✍️ Usage in a UPSC answer

The Supreme Court's SUPACE portal, which uses machine learning to sift through millions of case documents and surface relevant precedents for judges in real time, exemplifies how AI can augment judicial efficiency without displacing the discretionary and interpretive authority that the Constitution vests in the courts.

Synonyms

statistical learningautomated learningalgorithmic learningAI learningpredictive modelling (applied sense)

Antonyms

rule-based programmingexpert systemshard-coded logicmanual classification

🌱 Word Family

machine learning (n), supervised learning (n), unsupervised learning (n), reinforcement learning (n), ML model (n phrase), training data (n phrase), deep learning (n, subfield)

🔡 Root

Latin machina = device, engine (from Greek mākhana = contrivance); Old English leornian = to acquire knowledge — a machine that learns

📜 Etymology

The phrase 'machine learning' was coined by IBM researcher Arthur Samuel in 1959 in a paper on a self-improving checkers-playing program. Samuel defined it as a field of study giving computers the ability to learn without being explicitly programmed. The term predates artificial intelligence as a specific subfield label and has remained in continuous use since.

🧠 Memory Hook

MACHINE LEARNING: Arthur Samuel 1959 — his checkers program 'learned' to beat him by playing thousands of games. Picture the machine at a chess board, losing at first, improving each game, eventually winning: that is ML. The key: no explicit rules, just patterns from data. Coin the connection: Samuel 1959 = ML was born.

📝 Seen in UPSC Question Papers

Prelims 2025 — Technology
Prelims 2025 — IT & Cyber
Prelims 2024 — IT & Cyber

Real UPSC previous-year questions whose text uses “Machine learning” — proof this word earns its place on your list.

🔬 Science & Technology 📖 Seen in: Emerging Technologies — AI, Quantum Co...

Tip: press Alt+S to hear pronunciation