Well-calibrated predictions from on-line compression models

It has been shown recently that transductive confidence machine (TCM) is automatically well-calibrated when used in the on-line mode and provided that the data sequence is generated by an exchangeable distribution. In this paper we strengthen this result by relaxing the assumption of exchangeability of the data-generating distribution to the much weaker assumption that the data agrees with a given "on-line compression model".

[1]  A. Shiryaev,et al.  Probability (2nd ed.) , 1995, Technometrics.

[2]  S. S. Wilks Determination of Sample Sizes for Setting Tolerance Limits , 1941 .

[3]  David A. Freedman,et al.  Invariants Under Mixing which Generalize de Finetti's Theorem , 1962 .

[4]  Stephen C. Arnold,et al.  Kendall's advanced theory of statistics. Vol.2A: Classical inference and the linear model , 1999 .

[5]  Jammalamadaka Introduction to Linear Regression Analysis (3rd ed.) , 2003 .

[6]  Andrei N. Kolmogorov,et al.  Logical basis for information theory and probability theory , 1968, IEEE Trans. Inf. Theory.

[7]  S. Lauritzen Extremal Families and Systems of Sufficient Statistics , 1988 .

[8]  H. Seal Studies in the history of probability and statistics. XV. The historical velopment of the Gauss linear model. , 1967, Biometrika.

[9]  Vladimir Vovk,et al.  On-line confidence machines are well-calibrated , 2002, The 43rd Annual IEEE Symposium on Foundations of Computer Science, 2002. Proceedings..

[10]  P. Diaconis,et al.  Algebraic algorithms for sampling from conditional distributions , 1998 .

[11]  J. Brian Gray,et al.  Introduction to Linear Regression Analysis , 2002, Technometrics.

[12]  I. Guttman,et al.  Statistical Tolerance Regions: Classical and Bayesian , 1970 .

[13]  M. Kendall Theoretical Statistics , 1956, Nature.

[14]  Vladimir Vovk,et al.  Asymptotic Optimality of Transductive Confidence Machine , 2002, ALT.

[15]  H. Seal Studies in the history of probability and statistics , 1977 .

[16]  Vladimir Vovk,et al.  A Universal Well-Calibrated Algorithm for On-line Classification , 2004, J. Mach. Learn. Res..

[17]  Richard F. Gunst,et al.  Applied Regression Analysis , 1999, Technometrics.

[18]  Alexander Gammerman,et al.  Machine-Learning Applications of Algorithmic Randomness , 1999, ICML.

[19]  G. A. Baker The Probability That the Mean of a Second Sample Will Differ from the Mean of a First Sample By Less Than a Certain Multiple of the Standard Deviation of the First Sample , 1935 .

[20]  D. West Introduction to Graph Theory , 1995 .

[21]  E. Asarin Some Properties of Kolmogorov $\Delta$-Random Finite Sequences , 1988 .

[22]  Alexander Gammerman,et al.  Transduction with Confidence and Credibility , 1999, IJCAI.

[23]  Alexander Gammerman,et al.  Testing Exchangeability On-Line , 2003, ICML.

[24]  Per Martin-Löf,et al.  The Definition of Random Sequences , 1966, Inf. Control..

[25]  Vladimir Vovk,et al.  Kolmogorov's Contributions to the Foundations of Probability , 2003, Probl. Inf. Transm..

[26]  A. N. Kolmogorov Combinatorial foundations of information theory and the calculus of probabilities , 1983 .

[27]  Lakhmi C. Jain,et al.  Introduction to Bayesian Networks , 2008 .

[28]  V. Vovk On the concept of the Bernoulli property , 1986 .

[29]  Vladimir Vovk,et al.  Online Region Prediction with Real Teachers , 2003 .

[30]  Ilia Nouretdinov,et al.  Transductive Confidence Machine Is Universal , 2003, ALT.

[31]  Vladimir Vovk,et al.  Criterion of calibration for transductive confidence machine with limited feedback , 2006, Theor. Comput. Sci..

[32]  M. Kendall,et al.  Kendall's advanced theory of statistics , 1995 .

[33]  Frank Harary,et al.  Graph Theory , 2016 .