REWARD VERSUS RISK IN UNCERTAIN INFERENCE: THEOREMS AND SIMULATIONS

Systems of logico-probabilistic (LP) reasoning characterize inference from conditional assertions that express high conditional probabilities. In this paper we investigate four prominent LP systems, the systems O, P , Z , and QC . These systems differ in the number of inferences they licence ( O ⊂ P ⊂ Z ⊂ QC) . LP systems that license more inferences enjoy the possible reward of deriving more true and informative conclusions, but with this possible reward comes the risk of drawing more false or uninformative conclusions. In the first part of the paper, we present the four systems and extend each of them by theorems that allow one to compute almost-tight lower-probability-bounds for the conclusion of an inference, given lower-probability-bounds for its premises. In the second part of the paper, we investigate by means of computer simulations which of the four systems provides the best balance of reward versus risk . Our results suggest that system Z offers the best balance.

[1]  Raymond Reiter,et al.  A Logic for Default Reasoning , 1987, Artif. Intell..

[2]  Dov M. Gabbay,et al.  Theoretical Foundations for Non-Monotonic Reasoning in Expert Systems , 1989, Logics and Models of Concurrent Systems.

[3]  Peter Urbach,et al.  Scientific Reasoning: The Bayesian Approach , 1989 .

[4]  Markus Werning,et al.  Human Conditional Reasoning Explained by Non-Monotonicity and Probability: An Evolutionary Account , 2008 .

[5]  David Poole,et al.  A Logical Framework for Default Reasoning , 1988, Artif. Intell..

[6]  Vann McGee,et al.  Conditional Probabilities and Compounds of Conditionals , 1989 .

[7]  Peter Gärdenfors,et al.  Nonmonotonic Inference Based on Expectations , 1994, Artif. Intell..

[8]  Sarit Kraus,et al.  Nonmonotonic Reasoning, Preferential Models and Cumulative Logics , 1990, Artif. Intell..

[9]  William Harper,et al.  Motivating Objective Bayesianism: From Empirical Constraints to Objective Probabilities , 2006 .

[10]  Robert C. Moore.  Semantical Considerations on Nonmonotonic Logic , 1985, IJCAI.

[11]  Hannes Leitgeb,et al.  Finitistic and Frequentistic Approximation of Probability Measures with or without σ-Additivity , 2008, Stud Logica.

[12]  H. Kyburg,et al.  Nomic probability and the foundations of induction , 1990 .

[13]  Robert Stalnaker.  Probability and Conditionals , 1970, Philosophy of Science.

[14]  Henry Ely Kyburg,et al.  The logical foundations of statistical inference , 1974 .

[15]  Gerhard Schurz,et al.  Tweety, or Why Probabilism and even Bayesianism Need Objective and Evidential Probabilities , 2012 .

[16]  William Harper.  Rational Belief Change, Popper Functions and Counterfactuals , 1975 .

[17]  Krister Segerberg.  Notes on conditional logic , 1989, Stud Logica.

[18]  Hannes Leitgeb,et al.  Inference on the Low Level , 2004 .

[19]  J. Pollock.  Cognitive Carpentry: A Blueprint for How to Build a Person , 1995 .

[20]  Renée Elio,et al.  What Should Default Reasoning be, by Default? , 1997, Comput. Intell..

[21]  Michael Strevens,et al.  Do Large Probabilities Explain Better? , 2000, Philosophy of Science.

[22]  Drew McDermott,et al.  Non-Monotonic Logic I , 1987, Artif. Intell..

[23]  Jeff B. Paris,et al.  In defense of the maximum entropy inference process , 1997, Int. J. Approx. Reason..

[24]  Jeff B. Paris,et al.  When Maximizing Entropy gives the Rational Closure , 2003, J. Log. Comput..

[25]  Simon Parsons,et al.  Propagating Probabilities in System P , 1998, FLAIRS.

[26]  John F. Horty,et al.  Skepticism and floating conclusions , 2002, Artif. Intell..

[27]  Donald Bamber,et al.  Entailment with Near Surety of Scaled Assertions of High Conditional Probability , 2000, J. Philos. Log..

[28]  David E Over,et al.  Conditionals and conditional probability. , 2003, Journal of experimental psychology. Learning, memory, and cognition.

[29]  James P. Delgrande,et al.  An Approach to Default Reasoning Based on a First-Order Conditional Logic: Revised Report , 1987, Artif. Intell..

[30]  Gerhard Brewka,et al.  Nonmonotonic Reasoning: Logical Foundations of Commonsense By Gerhard Brewka (Cambridge University Press, 1991) , 1991, SGAR.

[31]  Van Fraassen,et al.  Laws and symmetry , 1989 .

[32]  J. Bennett.  A Philosophical Guide to Conditionals , 2003 .

[33]  Judea Pearl,et al.  Qualitative Probabilities for Default Reasoning, Belief Revision, and Causal Modeling , 1996, Artif. Intell..

[34]  Ernest W. Adams,et al.  The logic of ‘Almost all’ , 1974, J. Philos. Log..

[35]  Thomas Lukasiewicz,et al.  Probabilistic Deduction with Conditional Constraints over Basic Events , 2011, KR.

[36]  J. B. Paris,et al.  The Uncertain Reasoner's Companion: Bibliography , 1995 .

[37]  Gerhard Schurz,et al.  What Is 'normal'? an Evolution-theoretic Foundation of Normic Laws and Their Relation to Statistical Normality. What Is 'normal'? an Evolution-theoretic Foundation of Normic Laws and Their Relation to Statistical Normality , 2022 .

[38]  Hans Reichenbach,et al.  The theory of probability , 1968 .

[39]  Angelo Gilio.  Precise Propagation of Upper and Lower Probability Bounds in System P , 2000 .

[40]  Rohit Parikh,et al.  PROBABILISTIC CONDITIONALS ARE ALMOST MONOTONIC , 2008, The Review of Symbolic Logic.

[41]  Daniel Lehmann,et al.  What does a Conditional Knowledge Base Entail? , 1989, Artif. Intell..

[42]  Ernest W. Adams,et al.  On the logic of high probability , 1986, Journal of Philosophical Logic.

[43]  David Makinson,et al.  Conditional Probability in the Light of Qualitative Belief Change , 2011, J. Philos. Log..

[44]  D. Lewis.  Probabilities of Conditionals and Conditional Probabilities , 1976 .

[45]  David Lindley,et al.  Logical foundations of probability , 1951 .

[46]  Joseph Y. Halpern,et al.  Representation Dependence in Probabilistic Inference , 1995, IJCAI.

[47]  John F. Horty,et al.  Defaults with Priorities , 2007, J. Philos. Log..

[48]  Edwin T. Jaynes.  Prior Probabilities , 2010, Encyclopedia of Machine Learning.

[49]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems , 1988 .

[50]  Joseph Y. Halpern.  Reasoning about uncertainty , 2003 .

[51]  James Hawthorne,et al.  The Quantitative/Qualitative Watershed for Rules of Uncertain Inference , 2007, Stud Logica.

[52]  David S. Touretzky,et al.  A Clash of Intuitions: The Current State of Nonmonotonic Multiple Inheritance Systems , 1987, IJCAI.

[53]  Dov M. Gabbay,et al.  Nonmonotonic reasoning and uncertain reasoning , 1994 .

[54]  Ernest W. Adams,et al.  A note on comparing probabilistic and modal logics of conditionals , 2008 .

[55]  Wolfgang Spohn,et al.  Ordinal Conditional Functions: A Dynamic Theory of Epistemic States , 1988 .

[56]  David E. Over,et al.  Evolution and the Psychology of Thinking: The Debate , 2004 .

[57]  O. Wilhelm,et al.  The meaning(s) of conditionals: conditional probabilities, mental models, and personal utilities. , 2003, Journal of experimental psychology. Learning, memory, and cognition.

[58]  Wolfgang Spohn,et al.  Stochastic independence, causal independence, and shieldability , 1980, J. Philos. Log..

[59]  David Poole,et al.  The Effect of Knowledge on Belief: Conditioning, Specificity and the Lottery Paradox in Default Reasoning , 1991, Artif. Intell..

[60]  David Makinson,et al.  Bridges from classical to nonmonotonic logic , 2005, Texts in computing.

[61]  Jeff B. Paris,et al.  O IS NOT ENOUGH , 2009, The Review of Symbolic Logic.

[62]  P. Suppes.  Probabilistic Inference and the Concept of Total Evidence , 1969 .

[63]  Angelo Gilio,et al.  Probabilistic Reasoning Under Coherence in System P , 2002, Annals of Mathematics and Artificial Intelligence.

[64]  E. W. Adams,et al.  The logic of conditionals , 1975 .

[65]  Ernest W. Adams,et al.  Four probability-preserving properties of inferences , 1996, J. Philos. Log..

[66]  J. Hawthorne,et al.  Degree-of-Belief and Degree-of-Support: Why Bayesians Need Both Notions , 2005 .

[67]  Gerd Gigerenzer,et al.  How to Improve Bayesian Reasoning Without Instruction: Frequency Formats , 1995 .

[68]  John L. Pollock,et al.  Justification and Defeat , 1994, Artif. Intell..

[69]  Gerhard Schurz,et al.  Probabilistic Semantics for Delgrande's Conditional Logic and a Counterexample to His Default Logic , 1998, Artif. Intell..

[70]  Gerhard Schurz,et al.  Probabilistic Default Logic Based on Irrelevance and Relevance Assumptions , 1997, ECSQARU-FAPR.

[71]  Gerhard Schurz,et al.  Non-Monotonic Reasoning from an Evolution-Theoretic Perspective: Ontic, Logical and Cognitive Foundations , 2005, Synthese.

[72]  James Hawthorne,et al.  On the logic of nonmonotonic conditionals and conditional probabilities , 1996, J. Philos. Log..