The Speed Prior: A New Simplicity Measure Yielding Near-Optimal Computable Predictions (Proc. COLT 2002, based on quant-ph/0011122, 2000)

9/17/2002


Click here to start


Table of Contents

The Speed Prior: A New Simplicity Measure Yielding Near-Optimal Computable Predictions (Proc. COLT 2002, based on quant-ph/0011122, 2000)

Foundations of inductive inference

Optimal Inductive Inference

Bayesmix: Sharp Loss Bounds M. Hutter (on my SNF grant; IJFCS / ECML / ICML 2001)

Universal Mix

Problem: Universal M(x) not recursive

Ressource Postulate

Fastest way of making data

FAST & universal search (II)

Postulate+ FAST= Speed Prior S

S through Algorithm GUESS

S-based inductive inference

Approximating S through AS

S-based inductive inference II

S-based Inference III

Example: discrete pseudorandom universe (skip this if you don’t care for applications)

Consequences for physics etc (ignore this if you don’t care for applications)

Summary

Author: J. Schmidhuber

Email: juergen@idsia.ch

Home Page: http://www.idsia.ch/~juergen