Resources

Tutorial Slides

• G. Ochoa and N. Veerapen. SearchMaps: Visualising and Exploiting the Global Structure of Computational Search Spaces. GECCO 2018.
• S. Verel and A. Liefooghe. Fitness Landscape Analysis and Algorithm Performance for Single- and Multi-objective Combinatorial Optimization. CEC 2017.
• S. Verel. Fitness landscape analysis for understanding and designing local search heuristics. 51st CREST Open Workshop, 2017.
• D. Whitley. Next generation genetic algorithms. GECCO 2017.
• M. Tomassini. Introduction to Complex Networks. GECCO 2016.

---

Source Code

• LON-NPP-demo.R script for LON models creation, visualisation and metric calculation. Number Partitioning Problem (NPP) instances are used as case study. By G. Ochoa in EvoCOP 2017. LON datasets are here provided to be used with the script:
  • – LON data with N = 10, k = 0.4 before the phase transition, single funnel structure.
  • – LON data with N = 10, k = 1.0 after the phase transition, multi-funnel structure.
  • – LON dataset containing instances of size 10, 15 and 20, and with values of k from 0.4 to 1.2 in steps of 0.1.
• Code in Python for an example Iterated Local Search code instrumented to record the LON data through sampling of multiple runs. Code in R to parse the sampled data and build a LON. By N. Veerapen
• Code in C++ to exhaustively extract LONs of the Quadratic Assignment Problem (QAP). By S. Verel in CEC 2010.
• Code in Java for efficient generation of LON graphs on non-neutral landscapes. By J. E. Fieldsend and featured in GECCO 2018.

---

Datasets

• Data from Mapping the global structure of TSP fitness landscapes.
• Data from Understanding Phase Transitions with LONs: Number Partitioning as a Case Study.
• Data from Tunnelling Crossover Networks for the Asymmetric TSP.
• Data from Additional Dimensions to the Study of Funnels in Combinatorial Landscapes.
• Data from Deconstructing the Big Valley Search Space Hypothesis.

---

Related Websites

• Kateto.net provides some excellent tutorials in network visualistion using R and Gephi.
• Visualcomplexity.com provides a stunning gallery of visualisations.
• CDV Lab Computational Design and Visualization at the University of Coimbra, Portugal, is a multidisciplinary research lab featuring amazing visualisation projetcs.

---