from complexity to intelligence
play

From Complexity to Intelligence Introduction to Inductive Reasoning - PowerPoint PPT Presentation

From Complexity to Intelligence Introduction to Inductive Reasoning and Proportional Analogy 16 novembre 2016 Pierre-Alexandre Murena PAGE 1 / 77 Licence de droits dusage Table of contents Reminder Inductive Reasoning Deduction and


  1. From Complexity to Intelligence Introduction to Inductive Reasoning and Proportional Analogy 16 novembre 2016 Pierre-Alexandre Murena PAGE 1 / 77 Licence de droits d’usage

  2. Table of contents Reminder Inductive Reasoning Deduction and Induction Philosophical treatment Solomonoff’s theory of induction Proportional Analogy Analogy reasoning Hofstadter’s Micro-world Analogy and MDL Conclusion 16 novembre 2016 Pierre-Alexandre Murena PAGE 2 / 77 Licence de droits d’usage

  3. Kolmogorov Complexity How do you define the Kolmogorov complexity of a string x ? 16 novembre 2016 Pierre-Alexandre Murena PAGE 3 / 77 Licence de droits d’usage

  4. Kolmogorov Complexity How do you define the Kolmogorov complexity of a string x ? C M ( x ) = min { l ( p ); p () = x } p ∈ P M 16 novembre 2016 Pierre-Alexandre Murena PAGE 3 / 77 Licence de droits d’usage

  5. Conditional Kolmogorov Complexity How do you define the Kolmogorov complexity of a string x conditionnaly to a string y ? 16 novembre 2016 Pierre-Alexandre Murena PAGE 4 / 77 Licence de droits d’usage

  6. Conditional Kolmogorov Complexity How do you define the Kolmogorov complexity of a string x conditionnaly to a string y ? C M ( x | y ) = min { l ( p ); p ( y ) = x } p ∈ P M 16 novembre 2016 Pierre-Alexandre Murena PAGE 4 / 77 Licence de droits d’usage

  7. Minimum Description Length Principle What is the MDL Principle? 16 novembre 2016 Pierre-Alexandre Murena PAGE 5 / 77 Licence de droits d’usage

  8. Minimum Description Length Principle What is the MDL Principle? MDL Principle The best theory to describe observed data is the one which minimizes the sum of the description length (in bits) of : the theory description the data encoded from the theory 16 novembre 2016 Pierre-Alexandre Murena PAGE 5 / 77 Licence de droits d’usage

  9. Table of contents Reminder Inductive Reasoning Deduction and Induction Philosophical treatment Solomonoff’s theory of induction Proportional Analogy Analogy reasoning Hofstadter’s Micro-world Analogy and MDL Conclusion 16 novembre 2016 Pierre-Alexandre Murena PAGE 6 / 77 Licence de droits d’usage

  10. Table of contents Reminder Inductive Reasoning Deduction and Induction Philosophical treatment Solomonoff’s theory of induction Proportional Analogy Analogy reasoning Hofstadter’s Micro-world Analogy and MDL Conclusion 16 novembre 2016 Pierre-Alexandre Murena PAGE 7 / 77 Licence de droits d’usage

  11. Analysis of deduction Deduction examples (1) 1. All men are mortal. 2. Plato is a man. 3. Therefore, Plato is mortal. 16 novembre 2016 Pierre-Alexandre Murena PAGE 8 / 77 Licence de droits d’usage

  12. Analysis of deduction Deduction examples (2) Cauchy-Schwarz inequality Let α = ( a 1 , . . . , a n ) and β = ( b 1 , . . . , b n ) be two sequences of real numbers. Then : � n � � n � n � 2 � � a 2 � b 2 � ≥ a i b i i i i = 1 i = 1 i = 1 Proof 16 novembre 2016 Pierre-Alexandre Murena PAGE 9 / 77 Licence de droits d’usage

  13. Analysis of deduction Deduction examples (2) Cauchy-Schwarz inequality Let α = ( a 1 , . . . , a n ) and β = ( b 1 , . . . , b n ) be two sequences of real numbers. Then : � n � � n � n � 2 � � a 2 � b 2 � ≥ a i b i i i i = 1 i = 1 i = 1 Proof For any t ∈ R : 0 ≤ � α + t β � 2 = � α � 2 + 2 � α, β � t + � β � 2 t 2 = P ( t ) The quadratic polynomial P is positive, so its discriminant is negative : 4 |� α, β �| 2 − 4 � α � 2 � β � 2 ≤ 0 16 novembre 2016 Pierre-Alexandre Murena PAGE 9 / 77 Licence de droits d’usage

  14. Analysis of deduction Deduction examples (3) Strong perfect graph theorem A graph G is perfect if for every induced subgraph H , the chromatic number of H equals the size of the largest complete subgraph of H , and G is Berge if no induced subgraph of G is an odd cycle of length at least five or the complement of one. Proof 16 novembre 2016 Pierre-Alexandre Murena PAGE 10 / 77 Licence de droits d’usage

  15. Analysis of deduction Deduction examples (3) Strong perfect graph theorem A graph G is perfect if for every induced subgraph H , the chromatic number of H equals the size of the largest complete subgraph of H , and G is Berge if no induced subgraph of G is an odd cycle of length at least five or the complement of one. Proof 179 pages 16 novembre 2016 Pierre-Alexandre Murena PAGE 10 / 77 Licence de droits d’usage

  16. Analysis of deduction What is deduction? A definition for deductive reasoning Deductive reasoning is an approach where a set of logic rules are applied to general axioms in order to find (or more precisely to infer ) conclusions of no greater generality than the premises. 16 novembre 2016 Pierre-Alexandre Murena PAGE 11 / 77 Licence de droits d’usage

  17. Analysis of deduction What is deduction? A definition for deductive reasoning Deductive reasoning is an approach where a set of logic rules are applied to general axioms in order to find (or more precisely to infer ) conclusions of no greater generality than the premises. Or, less formally : General − → Less general General − → Particular 16 novembre 2016 Pierre-Alexandre Murena PAGE 11 / 77 Licence de droits d’usage

  18. Limits of deduction Will it rain today? 16 novembre 2016 Pierre-Alexandre Murena PAGE 12 / 77 Licence de droits d’usage

  19. Limits of deduction We are hardly able to get through one waking hour without facing some situation (e.g. will it rain or won’t it? ) where we do not have enough information to permit deductive reasoning; but still we must decide immediately. In spite of its familiarity, the formation of plausible conclusions is a very subtle process. in [Edwin T. Jaynes, Probability theory. The logic of science , Cambridge U. Press, 2003] 16 novembre 2016 Pierre-Alexandre Murena PAGE 13 / 77 Licence de droits d’usage

  20. Examples of conclusions of non-deductive reasoning It will rain today. All dogs bark. Everybody in this room knows that 1 + 1 = 2 The sun always rises in the East. Life is not a dream. . . . 16 novembre 2016 Pierre-Alexandre Murena PAGE 14 / 77 Licence de droits d’usage

  21. Inductive reasoning Definition Inductive reasoning is an approach in which the premises provide a strong evidence for the truth of the conclusion. The conclusion of induction is not guaranteed to be true! 16 novembre 2016 Pierre-Alexandre Murena PAGE 15 / 77 Licence de droits d’usage

  22. A frequent confusion Deduction : General rule = ⇒ Particular case Induction : Particular case = ⇒ General rule 16 novembre 2016 Pierre-Alexandre Murena PAGE 16 / 77 Licence de droits d’usage

  23. A frequent confusion Deduction : General rule = ⇒ Particular case Induction : Particular case = ⇒ General rule This is incorrect! 16 novembre 2016 Pierre-Alexandre Murena PAGE 16 / 77 Licence de droits d’usage

  24. Table of contents Reminder Inductive Reasoning Deduction and Induction Philosophical treatment Solomonoff’s theory of induction Proportional Analogy Analogy reasoning Hofstadter’s Micro-world Analogy and MDL Conclusion 16 novembre 2016 Pierre-Alexandre Murena PAGE 17 / 77 Licence de droits d’usage

  25. Philosophical treatment Epicurus (342-270 B.C.) Principle of Multiple Explanations : If more than one theory is consistent with the observations, keep all theories. 16 novembre 2016 Pierre-Alexandre Murena PAGE 18 / 77 Licence de droits d’usage

  26. Philosophical treatment Sextus Empiricus (160-210) When they propose to establish the universal from the particulars by means of induction, they will effect this by a review of either all or some of the particulars. But if they review some, the induction will be insecure, since some of the particulars omitted in the induction may contravene the universal ; while if they are to review all, they will be toiling at the impossible , since the particulars are infinite and indefinite. 16 novembre 2016 Pierre-Alexandre Murena PAGE 19 / 77 Licence de droits d’usage

  27. Philosophical treatment Sextus Empiricus (160-210) When they propose to establish the universal from the particulars by means of induction, they will effect this by a review of either all or some of the particulars. But if they review some, the induction will be insecure, since some of the particulars omitted in the induction may contravene the universal ; while if they are to review all, they will be toiling at the impossible , since the particulars are infinite and indefinite. 1. It is impossible to explore all possible situations. 2. How is it possible to know that the chosen individuals are representative of the concept? 16 novembre 2016 Pierre-Alexandre Murena PAGE 19 / 77 Licence de droits d’usage

  28. Philosophical treatment Example of a wrong induction Do birds fly? 16 novembre 2016 Pierre-Alexandre Murena PAGE 20 / 77 Licence de droits d’usage

  29. Philosophical treatment Example of a wrong induction Do birds fly? No! 16 novembre 2016 Pierre-Alexandre Murena PAGE 20 / 77 Licence de droits d’usage

  30. Philosophical treatment William of Ockham (1290-1349) Occam’s Razor Principle : Entities should not be multiplied beyond necessity 16 novembre 2016 Pierre-Alexandre Murena PAGE 21 / 77 Licence de droits d’usage

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend