Abstract
AbstractWe combine Solomonoff’s approach to universal prediction with algorithmic statistics and suggest to use the computable measure that provides the best “explanation” for the observed data (in the sense of algorithmic statistics) for prediction. In this way we keep the expected sum of squares of prediction errors bounded (as it was for the Solomonoff’s predictor) and, moreover, guarantee that the sum of squares of prediction errors is bounded along any Martin-Löf random sequence. An extended abstract of this paper was presented at the 16th International Computer Science Symposium in Russia (CSR 2021) (Milovanov 2021).
Funder
HORIZON EUROPE European Research Council
Unidade Local de Saúde de São José
Publisher
Springer Science and Business Media LLC
Reference15 articles.
1. Gács, P., Tromp, J., Vitányi, P.M.B.: Algorithmic statistics. IEEE Trans. Inf. Theory 47(6), 2443–2463 (2001)
2. Hutter, M., Poland, J.: Asymptotics of Discrete MDL for Online Prediction. IEEE Trans. Inf. Theory 51(11), 3780–3795 (2005)
3. Hutter, M.: Discrete MDL Predicts in Total Variation Advances in Neural Information Processing Systems 22 (NIPS–2009) 817-825
4. Hutter, M.: Sequential predictions based on algorithmic complexity. J. Comput. Syst. Sci. 72, 95–117 (2006)
5. LNCS (LNAI);M Hutter,2004