Information AgeScience & Math Backpropagation 1986Major Rumelhart, Hinton and Williams popularize a method to train multi-layer neural networks. ▲ MattersSaveShareTrace roots Longest chain: The Big Bang → First stars ignite → Supernovae seed heavy elements → The Sun forms → Earth forms → First oceans → First life → Photosynthesis evolves → Great Oxidation Event → Complex cells (eukaryotes) → Multicellular life → First animals → Cambrian Explosion → First vertebrates → Jawed fish → Fish walk onto land → The amniotic egg → The Great Dying → Dinosaurs appear → Asteroid ends the dinosaurs → Early primates → Apes split from monkeys → Human line splits from chimpanzees → Walking upright → First stone tools → Genus Homo appears → Homo erectus → Control of fire → Habitual fire and cooking → Homo sapiens appears → Complex language → Natufian villages → Oldest known bread → Agriculture begins → Irrigation → Uruk and the first cities → Cylinder seals → Writing (cuneiform) → Babylonian mathematics → Babylonian astronomical diaries → Thales and natural philosophy → Trial and death of Socrates → Plato's Academy → Aristotle's Lyceum → Alexander's conquests → Alexandria founded → Library of Alexandria → House of Wisdom → Al-Khwarizmi and algebra → Descartes and analytic geometry → Calculus → Binary arithmetic → Boolean logic → Gödel's incompleteness theorems → Turing machine → Perceptron → Backpropagation Community notesLoading notes… Learn more ranked by votesDocumentnature.comLearning representations by back-propagating errors, Rumelhart, Hinton and Williams 1986 (Nature) ↗Editor's pickBookTerrence J. SejnowskiThe Deep Learning Revolution ↗Editor's pickBookCade MetzGenius Makers ↗Editor's pickVideoSearch on YouTubebackpropagation explained neural networks ↗Editor's pick