✳flâneur — a map of the web's best reading

The Problem of Induction and Machine Learning | Vaden Masrani

vmasrani.github.io · 8,423 words · saved by 1 readers

Because of my fight against induction I have sometimes been described as an irrationalist. But if it is true that there is no induction, then it is irrational to continue to have faith in induction. … Who is more rational? The believer in generalizing induction, or the critic who tries hard to anticipate, and to eliminate, the mistakes that may lurk in his conjectures - and who warns the scientist that his admirable and growing scientific knowledge remains uncertain conjectural knowledge? Is it irrational to teach the distinction between truth and certainty? To say that we ought to search for truth, which we may discover, though we may never be certain of it? - K.P. The Non-Existence of Probabilistic Inductive Support, p. 316 A response to Induction, Popper, and Machine Learning. Bruce Nielson and Daniel C. Elton have recently uploaded a paper to the arXiv titled Induction, Popper, and Machine Learning (hereafter abbreviated IPML), which just so happen to be three subjects I’m intensel

The Problem of Induction and Machine Learning | Vaden Masrani The Problem of Induction and Machine Learning An exchange with Bruce Nielson Because of my fight against induction I have sometimes been described as an irrationalist. But if it is true that there is no induction, then it is irrational to continue to have faith in induction. … Who is more rational? The believer in generalizing induction, or the critic who tries hard to anticipate, and to eliminate, the mistakes that may lurk in his conjectures - and who warns the scientist that his admirable and growing scientific knowledge remains

Explore this link on the map →

saved by

related reading