a sequence based selection hyper heuristic utilising a
play

A Sequence-based Selection Hyper-heuristic Utilising a Hidden Markov - PowerPoint PPT Presentation

Introduction Proposed Method Case Studies Summary A Sequence-based Selection Hyper-heuristic Utilising a Hidden Markov Model Ahmed Kheiri Ed Keedwell College of Engineering, Mathematics and Physical Sciences The 43rd CREST Open Workshop -


  1. Introduction Proposed Method Case Studies Summary A Sequence-based Selection Hyper-heuristic Utilising a Hidden Markov Model Ahmed Kheiri Ed Keedwell College of Engineering, Mathematics and Physical Sciences The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  2. Introduction Proposed Method Case Studies Summary Outline Introduction Proposed Method Case Studies Summary The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  3. Introduction Proposed Method Case Studies Summary Outline Introduction Proposed Method Case Studies Summary The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  4. Introduction Proposed Method Case Studies Summary Search and Optimisation Search and optimisation algorithms are concerned with the discovery of the best possible solution in a given time to maximise or minimise an objective (or set of objectives). Most real-world search and optimisation problems cannot be solved exactly, requiring heuristic approaches such as LSs. Hyper-heuristics aim to automate the search process. — Burke et al., (2013) for recent survey Edmund K. Burke, Michel Gendreau, Matthew Hyde, Graham Kendall, Gabriela Ochoa, Ender Özcan and Rong Qu Hyper-heuristics: a survey of the state of the art. Journal of the Operational Research Society, 64(12):1695-1724, 2013. The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  5. Introduction Proposed Method Case Studies Summary Hyper-heuristic “A search method or learning mechanism for selecting or generating heuristics to solve computational search problems” Classification of approaches � � Select or generate heuristic � � Apply heuristic to solution Methodologies to select � � Accept or reject solution � � Analyse performance of low level heuristics � � Learn the selection mechanism Methodologies to generate ... Hyper-heuristic Online learning Offline learning Domain Barrier Problem Domain � � Representation Space of heuristics No learning � � Objective function � � Read/write Construction low level heuristics Instance Space of solutions � � Construct initial solution Perturbation low level heuristics ... The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  6. Introduction Proposed Method Case Studies Summary Selection Hyper-heuristic Framework return best obtained solution Domain barrier Selection hyper-heuristic yes evaluation function, Problem specific initial solution, instances, … information, Terminate? no Selection Method Move Acceptance Accept? Select LLH no yes Space of Heuristics Apply to S prev S prev ← S current LLH 1 LLH 2 LLH 3 LLH n S current Space of Solutions Update system parameters according to problem independent information and the performance of applied LLHs, … The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  7. Introduction Proposed Method Case Studies Summary Outline Introduction Proposed Method Case Studies Summary The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  8. Introduction Proposed Method Case Studies Summary Sequence-based Selection Hyper-heuristic Key feature: Sequence-based selection hyper-heuristics aim to analyse the performance of, and construct, sequences of heuristics. The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  9. Introduction Proposed Method Case Studies Summary Fitness Landscape and Low Level Heuristics Fitness landscape Objective Optimal The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  10. Introduction Proposed Method Case Studies Summary Fitness Landscape and Low Level Heuristics Apply a single low level heuristic Objective Current Optimal The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  11. Introduction Proposed Method Case Studies Summary Fitness Landscape and Low Level Heuristics Apply a single low level heuristic Objective Current Optimal The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  12. Introduction Proposed Method Case Studies Summary Fitness Landscape and Low Level Heuristics Apply a sequence of two low level heuristics Objective Current Optimal The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  13. Introduction Proposed Method Case Studies Summary Fitness Landscape and Low Level Heuristics Apply a sequence of three or more low level heuristics Objective Current Optimal The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  14. Introduction Proposed Method Case Studies Summary Sequence-based Selection Hyper-heuristic Framework Sequence-based Hyper-heuristic Problem Domain Traditional Hyper-heuristic S current ← S initial Construct solution LLH i LLH 1 Choose a low level heuristic Construct a sequence Add to a sequence Choose a low level heuristic SEQ.add(LLH i ) LLH n no sequence constructed? Apply selected heuristic yes Apply selected sequence S new ← apply(SEQ, S current ) Maintain S best Accept ( S current , S new ) clear(SEQ) S current Domain Barrier no terminate? yes return ( S best ) The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  15. Introduction Proposed Method Case Studies Summary Hidden Markov Model (HMM) ����������������� �� �� �� �� ���� �� ��� ��� ��� �� ��� ��� ��� �� ��� ��� ��� ����������� ������������������ �� �� �� �� �� ��� ��� ��� ��� �� ��� ��� ��� ��� �� ��� ��� ��� ��� The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  16. Introduction Proposed Method Case Studies Summary Sequence-based Hyper-heuristic Utilising HMM The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  17. Introduction Proposed Method Case Studies Summary Sequence-based Hyper-heuristic Utilising HMM The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  18. Introduction Proposed Method Case Studies Summary Sequence-based Hyper-heuristic Utilising HMM ����������������� ��������� ����������������� ��������� � � � � ���� ���� ���� ���� ���� ���� �� �� ���� ���� ��� ��� ��� ��� ��� ��� ��� ��� ��� ��� ���� ���� ���� ���� ���� ��� ��� ��� ���� ��� ��� ���� ��� ��� ��� ���� ��� ��� ���� ��� ��� ��� ���� ��� ��� ���� ��� ��� ��� ���� ��� ��� �������������� ���� ���� ��� � ������� ���� ���� ��� � � �������� ���� ��� �������� ���� ������� � � �������� �����! ��"����� ���� ����� The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

  19. Introduction Proposed Method Case Studies Summary Sequence-based Hyper-heuristic Utilising HMM ����������������� ��������� ����������������� ��������� � � � � ���� ���� ���� ���� ���� ���� �� �� ���� ���� ��� ��� ��� ��� ��� ��� ��� ��� ��� ��� ���� ���� ���� ���� ���� ��� ��� ��� ���� ��� ��� ���� ��� ��� ��� ���� ��� ��� ���� ��� ��� ��� ���� ��� ��� ���� ��� ��� ��� ���� ��� ��� �������������� ���� ���� ���� ��� � ������� ���� ���� ���� ��� � � �������� ���� ��� �������� ���� ���� � � ������� � � ������ �!�����" ��#����� � � ���$������� ���� ��$����" �������� ���� ���� ����� ����� The 43rd CREST Open Workshop - Hyper-Heuristics for Software Engineering University of Exeter

Recommend


More recommend