×

Bayesian inference for the weights in logarithmic pooling. (English) Zbl 07810179

Summary: Combining distributions is an important issue in decision theory and Bayesian inference. Logarithmic pooling is a popular method to aggregate expert opinions by using a set of weights that reflect the reliability of each information source. However, the resulting pooled distribution depends heavily on set of weights given to each opinion/prior and thus careful consideration must be given to the choice of weights. In this paper we review and extend the statistical theory of logarithmic pooling, focusing on the assignment of the weights using a hierarchical prior distribution. We explore several statistical applications, such as the estimation of survival probabilities, meta-analysis and Bayesian melding of deterministic models of population growth and epidemics. We show that it is possible learn the weights from data, although identifiability issues may arise for some configurations of priors and data. Furthermore, we show how the hierarchical approach leads to posterior distributions that are able to accommodate prior-data conflict in complex models.

MSC:

62-XX Statistics
60K35 Interacting random processes; statistical mechanics type models; percolation theory

Software:

Stan; GitHub; Ternary

References:

[1] Abbas, A. E. (2009). “A Kullback-Leibler view of linear and log-linear pools.” Decision Analysis, 6(1): 25-37.
[2] Aitchison, J. and Shen, S. M. (1980). “Logistic-normal distributions: Some properties and uses.” Biometrika, 67(2): 261-272. · Zbl 0433.62012 · doi:10.2307/2335470
[3] Alkema, L., Raftery, A. E., and Brown, T. (2008). “Bayesian melding for estimating uncertainty in national HIV prevalence estimates.” Sexually Transmitted Infections, 84(Suppl 1): i11-i16.
[4] Alkema, L., Raftery, A. E., Clark, S. J., et al. (2007). “Probabilistic projections of HIV prevalence using Bayesian melding.” The Annals of Applied Statistics, 1(1): 229-248. · Zbl 1129.62098 · doi:10.1214/07-AOAS111
[5] Anon. (1978). “Influenza in a Boarding School.” The British Medical Journal, 1: 587.
[6] Bagnoli, M. and Bergstrom, T. (2005). “Log-concave probability and its applications.” Economic Theory, 26(2): 445-469. · Zbl 1077.60012 · doi:10.1007/s00199-004-0514-4
[7] Berger, J. (2006). “The case for objective Bayesian analysis.” Bayesian Analysis, 1(3): 385-402. · Zbl 1331.62042 · doi:10.1214/06-BA115
[8] Biggerstaff, M., Cauchemez, S., Reed, C., Gambhir, M., and Finelli, L. (2014). “Estimates of the reproduction number for seasonal, pandemic, and zoonotic influenza: a systematic review of the literature.” BMC Infectious Diseases, 14(1): 480.
[9] Bochkina, N. A. and Green, P. J. (2014). “The Bernstein-von Mises theorem and nonregular models.” The Annals of Statistics, 42(5): 1850-1878. · Zbl 1305.62112 · doi:10.1214/14-AOS1239
[10] Bousquet, N. (2008). “Diagnostics of prior-data agreement in applied Bayesian analysis.” Journal of Applied Statistics, 35(9): 1011-1029. · Zbl 1474.62078 · doi:10.1080/02664760802192981
[11] Carpenter, B., Gelman, A., Hoffman, M. D., Lee, D., Goodrich, B., Betancourt, M., Brubaker, M., Guo, J., Li, P., and Riddell, A. (2017). “Stan: A probabilistic programming language.” Journal of Statistical Software, 76(1).
[12] Carvalho, L. M., Villela, D. A. M., Coelho, F. C., and Bastos, L. S. (2022). “Supplementary Material of “Bayesian Inference for the Weights in Logarithmic Pooling”.” Bayesian Analysis. · doi:10.1214/22-BA1311SUPP
[13] Coelho, F. C. and Codeço, C. T. (2009). “Dynamic modeling of vaccinating behavior as a function of individual beliefs.” PLoS Comput. Biol., 5(7): e1000425. · doi:10.1371/journal.pcbi.1000425
[14] DasGupta, A. (2011). “The exponential family and statistical applications.” In Probability for Statistics and Machine Learning, 583-612. Springer. · Zbl 1233.62001 · doi:10.1007/978-1-4419-9634-3
[15] Diaconis, P. and Ylvisaker, D. (1979). “Conjugate priors for exponential families.” The Annals of Statistics, 269-281. · Zbl 0405.62011
[16] French, S. (1985). “Group consensus probability distributions: A critical survey in Bayesian statistics.” Bayesian Statistics, 2.
[17] Frühwirth-Schnatter, S., Celeux, G., and Robert, C. P. (2019). Handbook of Mixture Analysis. CRC Press. · Zbl 1419.62001
[18] Genest, C. (1984). “A characterization theorem for externally Bayesian groups.” Annals of Statistics, 12(3): 1100-1105. · Zbl 0541.62003 · doi:10.1214/aos/1176346726
[19] Genest, C., McConway, K. J., and Schervish, M. J. (1986). “Characterization of externally Bayesian pooling operators.” The Annals of Statistics, 487-501. · Zbl 0602.62005 · doi:10.1214/aos/1176349934
[20] Genest, C., Weerahandi, S., and Zidek, J. V. (1984). “Aggregating opinions through logarithmic pooling.” Theory and Decision, 17(1): 61-70. · Zbl 0541.90002 · doi:10.1007/BF00140056
[21] Genest, C. and Zidek, J. V. (1986). “Combining probability distributions: A critique and an annotated bibliography.” Statistical Science, 114-135. · Zbl 0587.62017
[22] Guardoni, G. L. (2002). “On irrelevance of alternatives and opinion pooling.” Brazilian Journal of Probability and Statistics, 87-98. · Zbl 1049.62002
[23] Jackson, C. H., Jit, M., Sharples, L. D., and De Angelis, D. (2015). “Calibration of complex models through Bayesian evidence synthesis: a demonstration and tutorial.” Medical Decision Making, 35(2): 148-161.
[24] Jaynes, E. T. (1957). “Information theory and statistical mechanics. II.” Physical Review, 108: 171-190. · Zbl 0084.43701
[25] Jombart, T., Frost, S., Nouvellet, P., Campbell, F., and Sudre, B. (2019). outbreaks: A Collection of Disease Outbreak Data. R package version 1.6.0. URL https://github.com/reconhub/outbreaks
[26] Leutbecher, M. and Palmer, T. N. (2008). “Ensemble forecasting.” Journal of Computational Physics, 227(7): 3515-3539. · Zbl 1132.86308 · doi:10.1016/j.jcp.2007.02.014
[27] Li, Z. S., Guo, J., Xiao, N.-C., and Huang, W. (2017). “Multiple priors integration for reliability estimation using the Bayesian melding method.” In Reliability and Maintainability Symposium (RAMS), 2017 Annual, 1-6. IEEE.
[28] Lind, N. C. and Nowak, A. S. (1988). “Pooling expert opinions on probability distributions.” Journal of Engineering Mechanics, 114(2): 328-341.
[29] Lindley, D. V. (2013). Understanding Uncertainty. John Wiley & Sons. · doi:10.1002/9781118650158.indsp2
[30] McAlinn, K., Aastveit, K. A., Nakajima, J., and West, M. (2019). “Multivariate Bayesian predictive synthesis in macroeconomic forecasting.” Journal of the American Statistical Association. arXiv:1711.01667. Published online: Oct 9 2019. · Zbl 1441.91057 · doi:10.1080/01621459.2019.1660171
[31] McAlinn, K., Aastveit, K. A., and West, M. (2018). “Bayesian predictive synthesis – discussion of: Using stacking to average Bayesian predictive distributions, by Y. Yao et al.” Bayesian Analysis, 13: 971-973. · Zbl 1407.62090 · doi:10.1214/17-BA1091
[32] McAlinn, K. and West, M. (2019). “Dynamic Bayesian predictive synthesis in time series forecasting.” Journal of Econometrics, 210: 155-169. arXiv:1601.07463. · Zbl 1452.62697 · doi:10.1016/j.jeconom.2018.11.010
[33] Murray, J. D. (2002). Mathematical Biology I. An Introduction, volume 17 of Interdisciplinary Applied Mathematics. New York: Springer, 3rd edn. · Zbl 1006.92001
[34] Myung, I. J., Ramamoorti, S., and Bailey Jr, A. D. (1996). “Maximum entropy aggregation of expert predictions.” Management Science, 42(10): 1420-1436. · Zbl 0884.90005
[35] Neal, R. M. (2003). “Slice sampling.” The Annals of Statistics, 31(3): 705-767. · Zbl 1051.65007 · doi:10.1214/aos/1056562461
[36] Neuenschwander, B., Branson, M., and Spiegelhalter, D. J. (2009). “A note on the power prior.” Statistics in Medicine, 28(28): 3562-3566.
[37] Pennock, D. M. and Wellman, M. P. (1997). “Representing aggregate belief through the competitive equilibrium of a securities market.” In Geiger, D. and Shenoy, P. P. (eds.), Proceedings of the Thirteenth Conference on Uncertainty in Artificial Intelligence, 392-400. Morgan Kaufmann Publishers Inc.
[38] Poole, D. and Raftery, A. E. (2000). “Inference for deterministic simulation models: The Bayesian melding approach.” Journal of the American Statistical Association, 95(452): 1244-1255. · Zbl 1072.62544 · doi:10.1080/01621459.2000.10474324
[39] Raftery, A. E., Newton, M. A., Satagopan, J. M., and Krivitsky, P. N. (2007). “Estimating the integrated likelihood via posterior simulation using the harmonic mean identity.” In Bernardo, J. M., Bayarri, M. J., Berger, J. O., Dawid, A. P., Heckerman, D., Smith, A. F. M., and West, M. (eds.), Bayesian Statistics, 1-45. Oxford University Press.
[40] Robert, C. (2007). The Bayesian Choice: From Decision-Theoretic Foundations to Computational Implementation. Springer Science & Business Media. · Zbl 1129.62003
[41] Rufo, M., Martin, J., Pérez, C., et al. (2012a). “Log-linear pool to combine prior distributions: a suggestion for a calibration-based approach.” Bayesian Analysis, 7(2): 411-438. · Zbl 1330.62057 · doi:10.1214/12-BA714
[42] Rufo, M. J., Pérez, C. J., Martín, J., et al. (2012b). “A Bayesian approach to aggregate experts’ initial information.” Electronic Journal of Statistics, 6: 2362-2382. · Zbl 1295.62095 · doi:10.1214/12-EJS752
[43] Saumard, A. and Wellner, J. A. (2014). “Log-concavity and strong log-concavity: a review.” Statistics Surveys, 8: 45. · Zbl 1360.62055 · doi:10.1214/14-SS107
[44] Savchuk, V. P. and Martz, H. F. (1994). “Bayes Reliability Estimation Using Multiple Sources of Prior Information: Binomial Sampling.” IEEE Transactions on Reliability, 43(1): 138-144.
[45] Smith, M. R. (2017). “Ternary: An R Package for Creating Ternary Plots.” Zenodo.
[46] Talts, S., Betancourt, M., Simpson, D., Vehtari, A., and Gelman, A. (2018). “Validating Bayesian inference algorithms with simulation-based calibration.” arXiv preprint arXiv:1804.06788.
[47] West, M. (1984). “Bayesian aggregation.” Journal of the Royal Statistical Society. Series A (General), 600-607. · Zbl 0581.62004 · doi:10.2307/2981847
[48] Yao, Y., Vehtari, A., Simpson, D., Gelman, A., et al. (2018). “Using stacking to average Bayesian predictive distributions (with discussion).” Bayesian Analysis, 13(3): 917-1003. · Zbl 1407.62090 · doi:10.1214/17-BA1091
[49] Zhong, M., Goddard, N., and Sutton, C. (2015). “Latent Bayesian melding for integrating individual and population models.” In Advances in Neural Information Processing Systems, 3618-3626.
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.