Bias regularization in neural network models for general insurance pricing

  • Mario V. WüthrichEmail author
Original Research Paper


Generalized linear models have the important property of providing unbiased estimates on a portfolio level. This implies that generalized linear models manage to provide accurate prices on a portfolio level. On the other hand, neural networks may provide very accurate prices on an individual policy level, but state-of-the-art use of neural networks does not pay any attention to unbiasedness on the portfolio level. This is an implicit consequence of applying early stopping rules in gradient descent methods for model fitting. In the present paper we discuss this deficiency and we provide two different techniques to overcome this drawback of neural network model fitting.


Generalized linear model Exponential dispersion family Neural network Gradient descent method Unbiasedness Balance property Regression tree 



  1. 1.
    Bailey RA (1963) Insurance rates with minimum bias. Proc CAS 50:4–11Google Scholar
  2. 2.
    Barndorff-Nielsen O (2014) Information and exponential families: in statistical theory. Wiley, HobokenCrossRefGoogle Scholar
  3. 3.
    Breiman L, Friedman JH, Olshen RA, Stone CJ (1984) Classification and regression trees. Wadsworth Statistics/Probability SeriesGoogle Scholar
  4. 4.
    Bühlmann H (1980) An economic premium principle. ASTIN Bull 11(1):52–60MathSciNetCrossRefGoogle Scholar
  5. 5.
    Bühlmann H, Gisler A (2005) A course in credibility theory and its applications. Springer, BerlinzbMATHGoogle Scholar
  6. 6.
    Denuit M, Maréchal X, Pitrebois S, Walhin J-F (2007) Actuarial modelling of claim counts. Wiley, HobokenCrossRefGoogle Scholar
  7. 7.
    Draxler F, Veschgini K, Salmhofer M, Hamprecht FA (2019) Essentially no barriers in neural network energy landscape. arXiv:1803.00885v5
  8. 8.
    Esscher F (1932) On the probability function in the collective theory of risk. Scand Actuar J 1932(3):175–195CrossRefGoogle Scholar
  9. 9.
    Frees EW (2010) Regression modeling with actuarial and financial applications. Cambridge University Press, CambridgezbMATHGoogle Scholar
  10. 10.
    Garipov T, Izmailov P, Podoprikhin D, Vetrov D, Wilson AG (2018) Loss surfaces, mode connectivity, and fast ensembling of DNNs. arXiv:1802.10026v4
  11. 11.
    Goodfellow I, Bengio Y, Courville A (2016) Deep learning. MIT Press, CambridgezbMATHGoogle Scholar
  12. 12.
    Jørgensen B (1987) Exponential dispersion models. J R Stat Soc Ser B (Methodological) 49/2:127–145MathSciNetzbMATHGoogle Scholar
  13. 13.
    Jung J (1968) On automobile insurance ratemaking. ASTIN Bull 5:41–48CrossRefGoogle Scholar
  14. 14.
    McCullagh P, Nelder JA (1983) Generalized linear models. Chapman & Hall, LondonCrossRefGoogle Scholar
  15. 15.
    Nelder JA, Wedderburn RWM (1972) Generalized linear models. J R Stat Soc Ser A (General) 135/3:370–384CrossRefGoogle Scholar
  16. 16.
    Ohlsson E, Johansson B (2010) Non-life insurance pricing with generalized linear models. Springer, BerlinCrossRefGoogle Scholar
  17. 17.
    Wüthrich MV, Buser C (2016) Data analytics for non-life insurance pricing. SSRN Manuscript ID 2870308, Version of June 4, 2019Google Scholar

Copyright information

© EAJ Association 2019

Authors and Affiliations

  1. 1.RiskLabETH ZurichZurichSwitzerland

Personalised recommendations