♣ 23 ♣Learning Machines
Machine learning is not a new, nor a recent concept. The term “machine learning” (ML) was coined by Artur Samuel 1959 while he worked for IBM. A first book about machine learning for pattern recognition was “Learning Machines” – Nilsson (1965). Machine learning remained an active research topic, and the author was in the 1980s studying perceptron neural networks at the university.
Already in 1965, the first framework for “multilayer perceptrons” was published by Alexey Ivakhnenko and Lapa. Since 1986, we call those algorithms “deep learning,” thanks to Rina Dechter. However, it was only in 2015, when Google’s DeepMind was able to beat the best human Go player in the world with their AlpahGo system, that the term “deep learning” very popular and the old research field got a lot of new interest.
In the early days of machine learning (ML), one focussed on creating algorithms that could mimic human brain function, and achieve a form of artificial intelligence (AI). In the 1990s, computers were fast enough to make practical applications possible and the research became more focussed on wider applications, rather than focusing on reproducing intelligence artificially.
Rule based intelligence (as in the computer language Lisp), became less popular as a research subject and in general the focus shifted away from symbolic approaches – inherited from AI, and it became more and more a specific discipline ...
Get The Big R-Book now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
