Logical Approaches to Computational Barriers

Finite Prediction of Recursive Real-Valued Functions

Speaker:
| Eiju Hirowatari |

Author(s): |
Eiju Hirowatari, Kouichi Hirata and Tetsuhiro Miyahara |

Slot: |
Array, 11:30-11:50, col. 4 |

This paper concerns learning theory of {\em recursive real-valued functions\/} that are one of the formulations for the computable real function. Hirowatari et al. (2005) have introduced the finite prediction of recursive real-valued functions, which is based on a finite prediction machine that is a procedure to request finite examples of a recursive real-valued function f and a datum of a real number x, and to output a datum of a real number as the value of f(x). In this paper, we newly establish the interaction of the criterion RealFP for finite prediction of recursive real-valued functions and the criteria RealEx, RealCons, RealFin and RealNum! for inductive inference of recursive real-valued functions.

websites: Arnold Beckmann | 2006-04-29 |