Abstract
Recently a new account to the problem of induction has been developed [1], based on a priori advantages of regret-weighted meta-induction (RW) in online learning [2]. The claimed a priori advantages seem to contradict the no free lunch (NFL) theorem, which asserts that relative to a state-uniform prior distribution (SUPD) over possible worlds all (non-clairvoyant) prediction methods have the same expected predictive success. In this paper we propose a solution to this problem based on four novel results:
-
RW enjoys free lunches, i.e., its predictive long-run success dominates that of other prediction strategies.
-
Yet the NFL theorem applies to online prediction tasks provided the prior distribution is a SUPD.
-
The SUPD is maximally induction-hostile and assigns a probability of zero to all possible worlds in which RW enjoys free lunches. This dissolves the apparent conflict with the NFL.
-
The a priori advantages of RW can be demonstrated even under the assumption of a SUPD. Further advantages become apparent when a frequency-uniform distribution is considered.
This work was supported by the DFG (Deutsche Forschungsgemeinschaft), SPP 1516. For valuable help we are indebted to Ronald Ortner.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Schurz, G.: The meta-inductivist’s winning strategy in the prediction game: a new approach to Hume’s problem. Philos. Sci. 75, 278–305 (2008)
Cesa-Bianchi, N., Lugosi, G.: Prediction, Learning, and Games. Cambridge University Press, Cambridge (2006)
Wolpert, D., Macready, W.: No free lunch theorems for search. Technical report SFI-TR-95-02-010, Santa Fe Institute (1995)
Wolpert, D.: On the connection between in-sample testing and generalization error. Complex Syst. 6, 47–94 (1992)
Schaffer, C.: A conservation law for generalization performance. In: Machine Learning (Proceedings of ICML 1994), pp. 259–265. Morgan Kaufmann, Burlington (1994)
Rao, R., Gordon, D., Spears, W.: For every generalization action, is there really an equal and opposite reaction? In: Machine Learning (Proceedings of ICML 1995), pp. 471–479. Morgan Kaufmann, Burlington (1994)
Giraud-Carrier, C., Provost, F.: Toward a justification of meta-learning: is the no free lunch theorem a show-stopper? In: Proceedings of the ICML-2005 Workshop on Meta-learning, pp. 12–19 (2006)
Wolpert, D.: The lack of a priori distinctions between learning algorithms. Neural Comput. 8, 1341–1390 (1996)
Schurz, G., Thorn, P.: The revenge of ecological rationality: strategy-selection by meta-induction. Minds Mach. 26(1), 31–59 (2016)
Schurz, G.: Meta-induction is an optimal prediction strategy. In: Proceedings of the 18th Annual Belgian-Dutch Conference on Machine Learning, pp. 66–74. University of Tilburg (2009)
Gigerenzer, G., Todd, P.M., the ABC Research Group. (eds.): Simple Heuristics That Make Us Smart. Oxford University Press, Dordrecht (1999)
Schurz, G.: The Optimality of Meta-induction. A New Approach to Hume’s Problem. Book manuscript (2017)
Carnap, R.: Logical Foundations of Probability, 2nd edn. University of Chicago Press, Chicago (1962)
Howson, C., Urbach, P.: Scientific Reasoning: The Bayesian Approach, 2nd edn. Open Court, Chicago (1996)
Solomonoff, R.: A formal theory of inductive inference. Inf. Control 7(1–22), 224–254 (1964)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Schurz, G., Thorn, P. (2017). A Priori Advantages of Meta-Induction and the No Free Lunch Theorem: A Contradiction?. In: Kern-Isberner, G., Fürnkranz, J., Thimm, M. (eds) KI 2017: Advances in Artificial Intelligence. KI 2017. Lecture Notes in Computer Science(), vol 10505. Springer, Cham. https://doi.org/10.1007/978-3-319-67190-1_18
Download citation
DOI: https://doi.org/10.1007/978-3-319-67190-1_18
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-67189-5
Online ISBN: 978-3-319-67190-1
eBook Packages: Computer ScienceComputer Science (R0)