CEC2015 Special Session

Special Session & Competition on Real-Parameter Single Objective Optimization (3 Different Cases)

at CEC-2015, Sendai International Centre, Sendai, Japan, 25-28 May 2015

If you face any difficulties, please inform me ( epnsugan@ntu.edu.sg ).

https://github.com/P-N-Suganthan

1. Call for papers

2. Q. Chen, B. Liu, Q. Zhang, J. J. Liang, P. N. Suganthan, B. Y. Qu, "Problem Definition and Evaluation Criteria for CEC 2015 Special Session and Competition on Bound Constrained Single-Objective Computationally Expensive Numerical Optimization", Technical Report, Computational Intelligence Laboratory, Zhengzhou University, Zhengzhou, China and Technical Report, Nanyang Technological University, Singapore, Nov 2014.

3. J. J. Liang, B. Y. Qu, P. N. Suganthan, Q. Chen, "Problem Definitions and Evaluation Criteria for the CEC 2015 Competition on Learning-based Real-Parameter Single Objective Optimization", Technical Report, Computational Intelligence Laboratory, Zhengzhou University, Zhengzhou, China and Technical Report, Nanyang Technological University, Singapore, Nov 2014.

4. B. Y. Qu, J. J. Liang, Z. Y. Wang, Q. Chen, P. N. Suganthan, "Novel Benchmark Functions for Continuous Multimodal Optimization with Comparative Results," Swarm and Evolutionary Computation, doi:10.1016/j.swevo.2015.07.003. [ <-- We request the usage of the new reference. Old Reference: B. Y. Qu, J. J. Liang, P. N. Suganthan, Q. Chen, "Problem Definitions and Evaluation Criteria for the CEC 2015 Competition on Single Objective Multi-Niche Optimization", Technical Report, Computational Intelligence Laboratory, Zhengzhou University, Zhengzhou, China and Technical Report, Nanyang Technological University, Singapore, Nov 2014. (Final Solution Submission Format is Given Here) ]

Comparison slides for all 3 cases are available in the respective folders: https://github.com/P-N-Suganthan

Learning Based Papers

Paper ID

Algorithm

Title

15031

SPS-L-SHADE-EIG

A Self-Optimization Approach for L-SHADE Incorporated with Eigenvector-Based Crossover and Successful-Parent-Selecting Framework on CEC 2015 Benchmark Set.

Rank - #1

15096

TEBO

Tuning Maturity Model of Ecogeography-Based Optimization On CEC 2015 Single-Objective Optimization Test Problems

15170

MVMO

Testing MVMO on Learning-based Real-Parameter Single Objective Benchmark Optimization Problems

Rank - #3

15230

LSHADE-ND

Neurodynamic Differential Evolution Algorithm and Solving CEC2015 Competition Problems

Rank - #3

15287

ICMLSP

An Improved Covariance Matrix Leaning and Searching Preference Algorithm for Solving CEC 2015 Benchmark Problems

15460

SaDPSO

A Self-adaptive Dynamic Particle Swarm Optimizer

15473

cooperation

Cooperation of Optimization Algorithms: A Simple Hierarchical Model

15485

hCC

Hybrid Cooperative Co-evolution For The CEC15 Benchmarks

15527

ABC-X-LS

A Configurable Generalized Artificial Bee Colony Algorithm with Local Search Strategies

15598

dynFWA

Dynamic Search Fireworks Algorithm for Solving CEC2015 Competition Problems

15620

DEsPA

A Differential Evolution Algorithm with Successbased Parameter Adaptation for CEC2015 Learning based Optimization

Rank - #2

15642

dynFWACM

Dynamic Search Fireworks Algorithm with Covariance Mutation for Solving the CEC 2015 Learning Based Competition Problems

15667

HumanCog

HumanCog: A Cognitive Architecture for Solving Optimization Problems

Expensive Optimization Papers

Paper ID

Algorithm

Score / Rank

E-15035

MVMO

3,062,550.15

E-15487

TunedCMAES

203,324,192.51

E-15664

CMAS-ES_QR

475,807,278.19

E-15667

iSRPSO

9,213,589,132.86

E-15682

humanCog

106,093,535,263.79

Multi-Niche optimization Papers

#15611: An evolutionary algorithm based on decomposition for multimodal optimization problems. (Rank 1)

#15568: Multimodal optimization using particle swarm optimization algorithms: CEC 2015 competition on single objective multi-niche optimization. (Rank 2)

#15629: Solving CEC 2015 Multimodal competition problems using neighborhood based speciation differential evolution. (Rank 3)

Paper Submission

Authors must strictly follow the manuscript preparation instructions: Available Here

When submitting, please make sure you select "SS04: Single Objective Numerical Optimization" as the "Main Research Topic" for all papers making use of these benchmarks. Each submission should have either the learning-based or expensive or multiple-niche problems (i.e. these 3 sets of problems should not be combined into a single paper).

Paper Submission Deadline

We will follow the extended deadlines as determined by CEC 2015. (Currently 19th of Dec 2014). If you're unable to meet the final extended deadline, please inform us ( epnsugan@ntu.edu.sg ).

Updates:

#2 on 7th Jan 2015

Niching Problems: (updated on 7th Jan 2015)

Codes:

1. Test function 5, there is a bug in the code and will affect the result. Three versions have been updated.

2. Test function 7, the accuracy is increased, the former "4.1265" in the function which is used to make f(x*)=0 is updated to "4.126514".

TR:

1. Errors in test functions 6, 7 and 8 are corrected.

2. pages are inserted.

#1 on 3rd Jan 2015

Learning-Based Problems：(updated on 3rd Jan 2015)

lamada definded for "Rotated Expanded Scaffer’s F6 Function" in the composition function is modified since the previous one is too small to make the whole subregion has a low function values. 5e-4 is modified to 10. Affected functions:

page 15: (12) Composition Function 4

page 16: (13) Composition Function 5

page 17: (14) Composition Function 6

page 18: (15) Composition Function 7

The codes are all updated and authors do not have to correct the results now, but for the final paper, they should rerun these four functions.

Niching Problems: (updated on 3rd Jan 2015)

1. 11)Composition Function 3 bias= [0, 10, 20, 30, 40, 50, 60, 70, 80, 90] is corrected to bias = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0] +F11*. The codes are correct. So, not to be modified.

2. lamada definded for "Rotated Expanded Scaffer’s F6 Function" in the composition function is modified to "100" since the previous one is too small to make the whole subregion to have low function values. Affected functions:

11) Composition Function 3

12) Composition Function 4

13) Composition Function 5

15) Composition Function 7

The codes are all updated and authors do not have to correct the results now, but for the final paper, they should rerun these four functions.

Request for Feedback

If you have suggestions to improve the technical report or if you find any potential bug in the codes, please inform us ( epnsugan@ntu.edu.sg ).