Monotone Conditional Complexity Bounds on Future Prediction Errors

14 years 1 months ago

Download www.idsia.ch

We bound the future loss when predicting any (computably) stochastic sequence online. Solomonoﬀ ﬁnitely bounded the total deviation of his universal predictor M from the true distribution µ by the algorithmic complexity

Alexey V. Chernov, Marcus Hutter

Real-time Traffic

ALT 2005 | Machine Learning | Stochastic Sequence Online | Total Deviation | Universal Predictor |

claim paper

» Algorithmic Complexity Bounds on Future Prediction Errors

» Approximate predictive state representations

» Nonparametric Time Series Prediction Through Adaptive Model Selection

» A Complete and Tight AverageCase Analysis of Learning Monomials

» Tractable Inference for Complex Stochastic Processes

» Action Reaction Learning Automatic Visual Analysis and Synthesis of Interactive Behaviour

» Probabilistic Inference Protection on Anonymized Data

» The Effects of the SubCarrier Grouping on MultiCarrier Channel Aware Scheduling

Post Info
More Details (n/a)

Added	14 Mar 2010
Updated	14 Mar 2010
Type	Conference
Year	2005
Where	ALT
Authors	Alexey V. Chernov, Marcus Hutter

Comments (0)

Sciweavers

Monotone Conditional Complexity Bounds on Future Prediction Errors

ALT 2005 | Machine Learning | Stochastic Sequence Online | Total Deviation | Universal Predictor |

Explore & Download

Productivity Tools

Document Tools

Image Tools

Sciweavers