On-Line Confidence Machines Are Well-Calibrated
FOCS '02 Proceedings of the 43rd Symposium on Foundations of Computer Science
Ridge Regression Confidence Machine
ICML '01 Proceedings of the Eighteenth International Conference on Machine Learning
Machine-Learning Applications of Algorithmic Randomness
ICML '99 Proceedings of the Sixteenth International Conference on Machine Learning
Transduction with Confidence and Credibility
IJCAI '99 Proceedings of the Sixteenth International Joint Conference on Artificial Intelligence
Hi-index | 0.00 |
This paper is concerned with the problem of on-line prediction in the situation where some data are unlabelled and can never be used for prediction, and even when the data are labelled, the labels may arrive with a delay. We construct a modification of randomised transductive confidence machine for this case and prove a necessary and sufficient condition for its predictions being calibrated, in the sense that in the long run they are wrong with a prespecified probability under the assumption that the data are generated independently by the same distribution. The condition for calibration turns out to be very weak: feedback should be given on more than a logarithmic fraction of steps.