Evolving Boolean Functions with Conjunctions and Disjunctions via Genetic Programming

被引:5
|
作者
Doerr, Benjamin [1 ]
Lissovoi, Andrei [2 ]
Oliveto, Pietro S. [2 ]
机构
[1] Ecole Polytech, CNRS, Lab Informat LIX, Palaiseau, France
[2] Univ Sheffield, Dept Comp Sci, Sheffield, S Yorkshire, England
基金
英国工程与自然科学研究理事会;
关键词
Theory; Genetic programming; Running time analysis;
D O I
10.1145/3321707.3321851
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Recently it has been proved that simple GP systems can efficiently evolve the conjunction of n variables if they are equipped with the minimal required components. In this paper, we make a considerable step forward by analysing the behaviour and performance of a GP system for evolving a Boolean function with unknown components, i.e. the target function may consist of both conjunctions and disjunctions. We rigorously prove that if the target function is the conjunction of n variables, then a GP system using the complete truth table to evaluate program quality evolves the exact target function in O(ln log(2) n) iterations in expectation, where l >= n is a limit on the size of any accepted tree. Additionally, we show that when a polynomial sample of possible inputs is used to evaluate solution quality, conjunctions with any polynomially small generalisation error can be evolved with probability 1 - O(log(2) (n)/n). To produce our results we introduce a super-multiplicative drift theorem that gives significantly stronger runtime bounds when the expected progress is only slightly super-linear in the distance from the optimum.
引用
收藏
页码:1003 / 1011
页数:9
相关论文
共 50 条
  • [1] On the Time and Space Complexity of Genetic Programming for Evolving Boolean Conjunctions
    Lissovoi, Andrei
    Oliveto, Pietro S.
    THIRTY-SECOND AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTIETH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE / EIGHTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2018, : 1363 - 1370
  • [2] On the Time and Space Complexity of Genetic Programming for Evolving Boolean Conjunctions
    Lissovoi, Andrei
    Oliveto, Pietro S.
    JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH, 2019, 66 : 655 - 689
  • [3] On the Analysis of Simple Genetic Programming for Evolving Boolean Functions
    Mambrini, Andrea
    Oliveto, Pietro S.
    GENETIC PROGRAMMING, EUROGP 2016, 2016, 9594 : 99 - 114
  • [4] (1+1) genetic programming with functionally complete instruction sets can evolve Boolean conjunctions and disjunctions with arbitrarily small error
    Doerr, Benjamin
    Lissovoi, Andrei
    Oliveto, Pietro S.
    ARTIFICIAL INTELLIGENCE, 2023, 319
  • [5] Discovering Non-Linear Boolean Functions by Evolving Walsh Transforms with Genetic Programming
    Rovito, Luigi
    De Lorenzo, Andrea
    Manzoni, Luca
    ALGORITHMS, 2023, 16 (11)
  • [6] Evolving kernel functions for SVMs by genetic programming
    Diosan, Laura
    Rogozan, Alexandrina
    Pecuchet, Jean-Pierre
    ICMLA 2007: SIXTH INTERNATIONAL CONFERENCE ON MACHINE LEARNING AND APPLICATIONS, PROCEEDINGS, 2007, : 19 - 24
  • [7] Evolving hash functions by means of genetic programming
    Estebanez, Cesar
    Cesar, Julio
    Ribagorda, Arturo
    GECCO 2006: GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE, VOL 1 AND 2, 2006, : 1861 - +
  • [8] Designing Bent Boolean Functions With Parallelized Linear Genetic Programming
    Husa, Jakub
    Dobai, Roland
    PROCEEDINGS OF THE 2017 GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE COMPANION (GECCO'17 COMPANION), 2017, : 1825 - 1832
  • [9] Comparison of Genetic Programming Methods on Design of Cryptographic Boolean Functions
    Husa, Jakub
    GENETIC PROGRAMMING, EUROGP 2019, 2019, 11451 : 228 - 244
  • [10] THE REALIZATION OF BOOLEAN FUNCTIONS BY DISJUNCTIONS OF PRODUCTS OF LINEAR-FORMS
    ALEKSANIAN, AA
    DOKLADY AKADEMII NAUK SSSR, 1989, 304 (04): : 781 - 784