Abstract
An extended version of the principle of empirical risk minimization is proposed. It is based on the application of averaging aggregation functions, rather than arithmetic means, to compute empirical risk. This is justified if the distribution of losses has outliers or is substantially distorted, which results in that the risk estimate becomes biased from the very beginning. In this case, for optimizing parameters, a robust estimate of the mean risk should be used. Such estimates can be constructed by using averaging aggregation functions, which are the solutions of the problem of minimizing the function of penalty for deviation from the mean value. An iterative reweighting scheme for numerically solving the problem of empirical risk minimization is proposed. Illustrative examples of the construction of a robust procedure for estimating parameters in the linear regression problem and in the problem of linearly separating two classes based on the application of an averaging mean function, which replaces the α-quantile, are given.
Article PDF
Similar content being viewed by others
Avoid common mistakes on your manuscript.
References
V. Vapnik, The Nature of Statistical Learning Theory (Springer-Verlag, Berlin, 2000).
Yu. I. Zhuravlev, in Problems of Applied Mathematics and Mechanics (Nauka, Moscow, 1971), pp. 67–74 [in Russian].
P. J. Huber, Robust Statistics (Wiley, New York, 1981; Mir, Moscow, 1984).
P. J. Rousseeuw, J. Am. Stat. Ass., No. 79, 871–880 (1984).
P. J. Rousseeuw and A. M. Leroy, Robust Regression and Outlier Detection (Wiley, New York, 1987).
Y. Ma, L. Li, X. Huang, and S. Wang, IFAC Proc. Vols, 44 (1), 11208–11213 (2011).
G. Beliakov, H. Sola, and T. Calvo, A Practical Guide to Averaging Functions (Springer-Verlag, Berlin, 2016).
T. Calvo and G. Beliakov, Fuzzy Sets Systems 161 (10), 1420–1436 (2010).
Z. M. Shibzukhov, Pat. Rec. Image Anal. 24 (3), 377–382 (2014).
Z. M. Shibzukhov, Dokl. Math. 91 (3), 391–393 (2015).
Author information
Authors and Affiliations
Corresponding author
Additional information
Original Russian Text © Z.M. Shibzukhov, 2017, published in Doklady Akademii Nauk, 2017, Vol. 476, No. 5, pp. 495–499.
Rights and permissions
About this article
Cite this article
Shibzukhov, Z.M. On the principle of empirical risk minimization based on averaging aggregation functions. Dokl. Math. 96, 494–497 (2017). https://doi.org/10.1134/S106456241705026X
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S106456241705026X