### Abstract

We consider the problem of sequential prediction of real-valued sequences using piece-wise linear models under the square-error loss function. In this context, we demonstrate a sequential algorithm for prediction whose accumulated squared error for every bounded sequence is asymptotically as small as that of the best fixed predictor for that sequence taken from the class of piecewise linear predictors. We also show that this predictor is optimal in certain settings in a particular min-max sense. This approach can also be applied to the class of piecewise constant predictors, for which a similar universal sequential algorithm can be derived with corresponding min-max optimality.

Original language | English (US) |
---|---|

Number of pages | 1 |

Journal | IEEE International Symposium on Information Theory - Proceedings |

State | Published - Oct 20 2004 |

Event | Proceedings - 2004 IEEE International Symposium on Information Theory - Chicago, IL, United States Duration: Jun 27 2004 → Jul 2 2004 |

### ASJC Scopus subject areas

- Theoretical Computer Science
- Information Systems
- Modeling and Simulation
- Applied Mathematics

## Fingerprint Dive into the research topics of 'Universal piecewise linear least squares prediction'. Together they form a unique fingerprint.

## Cite this

*IEEE International Symposium on Information Theory - Proceedings*.