Necessary and Sufficient Conditions for Optimal Decision Trees using Dynamic Programming

被引:0
|
作者
van der Linden, Jacobus G. M. [1 ]
de Weerdt, Mathijs M. [1 ]
Demirovic, Emir [1 ]
机构
[1] Delft Univ Technol, Dept Comp Sci, Delft, Netherlands
关键词
CLASSIFICATION TREES; KNOWLEDGE; STATE;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Global optimization of decision trees has shown to be promising in terms of accuracy, size, and consequently human comprehensibility. However, many of the methods used rely on general-purpose solvers for which scalability remains an issue. Dynamic programming methods have been shown to scale much better because they exploit the tree structure by solving subtrees as independent subproblems. However, this only works when an objective can be optimized separately for subtrees. We explore this relationship in detail and show the necessary and sufficient conditions for such separability and generalize previous dynamic programming approaches into a framework that can optimize any combination of separable objectives and constraints. Experiments on five application domains show the general applicability of this framework, while outperforming the scalability of general-purpose solvers by a large margin.
引用
收藏
页数:40
相关论文
共 50 条
  • [1] NECESSARY AND SUFFICIENT DYNAMIC PROGRAMMING CONDITIONS FOR CONTINUOUS TIME STOCHASTIC OPTIMAL CONTROL
    RISHEL, R
    SIAM JOURNAL ON CONTROL, 1970, 8 (04): : 559 - &
  • [2] NECESSARY AND SUFFICIENT DYNAMIC PROGRAMMING CONDITIONS FOR CONTINUOUS TIME STOCHASTIC OPTIMAL CONTROL
    RISHEL, RW
    NOTICES OF THE AMERICAN MATHEMATICAL SOCIETY, 1970, 17 (01): : 252 - &
  • [4] NECESSARY AND SUFFICIENT DYNAMIC-PROGRAMMING CONDITIONS FOR OPTIMAL-CONTROL PROBLEM WITH STATE CONSTRAINTS
    KHRUSTALEV, MM
    LECTURE NOTES IN CONTROL AND INFORMATION SCIENCES, 1990, 143 : 311 - 320
  • [5] Fair and Optimal Decision Trees: A Dynamic Programming Approach
    van der Linden, Jacobus G. M.
    de Weerdt, Mathijs M.
    Demirovic, Emir
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 35 (NEURIPS 2022), 2022,
  • [6] Necessary and sufficient conditions of some strong optimal solutions to the interval linear programming
    Li, Wei
    Luo, Jiajia
    Deng, Chongyang
    LINEAR ALGEBRA AND ITS APPLICATIONS, 2013, 439 (10) : 3241 - 3255
  • [7] MurTree: Optimal Decision Trees via Dynamic Programming and Search
    Demirovic, Emir
    Lukina, Anna
    Hebrard, Emmanuel
    Chan, Jeffrey
    Bailey, James
    Leckie, Christopher
    Ramamohanarao, Kotagiri
    Stuckey, Peter J.
    JOURNAL OF MACHINE LEARNING RESEARCH, 2022, 23
  • [8] MurTree: Optimal Decision Trees via Dynamic Programming and Search
    Demirović, Emir
    Lukina, Anna
    Hebrard, Emmanuel
    Chan, Jeffrey
    Bailey, James
    Leckie, Christopher
    Ramamohanarao, Kotagiri
    Stuckey, Peter J.
    Journal of Machine Learning Research, 2022, 23
  • [9] Learning optimal decision trees using constraint programming
    Verhaeghe, Helene
    Nijssen, Siegfried
    Pesant, Gilles
    Quimper, Claude-Guy
    Schaus, Pierre
    CONSTRAINTS, 2020, 25 (3-4) : 226 - 250
  • [10] Learning Optimal Decision Trees using Constraint Programming
    Verhaeghe, Helene
    Nijssen, Siegfried
    Pesant, Gilles
    Quimper, Claude-Guy
    Schaus, Pierre
    PROCEEDINGS OF THE TWENTY-NINTH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE, 2020, : 4765 - 4769