Section: STEM · Computer ScienceDifficulty: Medium

Machine Learning

USUK
US/məˈʃin ˈlərnɪŋ/UK/məˈʃin ˈlənɪŋ/

A subset of AI where systems learn from data to improve performance without explicit programming.

Also: ML

Definition

Machine learning is a branch of artificial intelligence in which algorithms and statistical models enable computer systems to learn from data and improve their performance on tasks without being explicitly programmed with rules. It encompasses supervised learning, unsupervised learning, and reinforcement learning. Applications include image recognition, natural language processing, recommendation systems, and fraud detection.

Example

“A spam email filter uses machine learning: it is trained on millions of labeled emails (spam and not spam), learning patterns in words, sender addresses, and formatting, then applies this knowledge to classify new incoming emails accurately.”

Usage Examples

  1. 1

    “The bank's machine learning model flagged the unusual transaction pattern within milliseconds, preventing a ₹5 lakh fraud.”

  2. 2

    “Using machine learning, the logistics company reduced delivery time predictions error from ±4 hours to ±45 minutes.”

When & How to Use

Use when enabling a system to improve its performance on a task through experience (data) without being explicitly programmed for every scenario.

  • ▸Training a fraud detection model on historical transaction data for a fintech company
  • ▸Building a recommendation engine that personalises product suggestions on an e-commerce platform
  • ▸Predicting employee attrition risk from HR data to enable proactive retention actions

Etymology & Origin

Coined by Arthur Samuel in 1959 in his paper on a checkers-playing program. 'Machine' from Latin 'machina'; 'learning' from Old English 'leornian' (to get knowledge).

History & Evolution

Early ML was dominated by symbolic AI (1950s–80s). Neural networks gained traction in the 1980s–90s. The deep learning revolution began with AlexNet winning ImageNet (2012). Today, transformer models (GPT, BERT) represent the frontier. India's IITs and startups now lead applied ML research.

Synonyms

  • ML
  • statistical learning
  • data-driven learning
  • predictive modeling

Antonyms / Opposites

  • rule-based programming
  • explicit programming

Images

CC-licensed · free to use
More on Wikimedia
Loading images…

Video

Dictionary Entry

Back to STEM