Fuzzy-Classification Assisted Solution Preselection in Evolutionary Optimization
Authors: Aimin Zhou, Jinyuan Zhang, Jianyong Sun, Guixu Zhang2403-2410
AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | The proposed FCPS scheme is applied to two state-of-the-art evolutionary algorithms on a test suite. The experimental results show the potential of FCPS on improving algorithm performance. Experimental Study Algorithms in Study The proposed FCPS strategy is integrated into two state-of-the-art evolutionary optimization algorithms, i.e., the composite differential evolution (Co DE) (Wang, Cai, and Zhang 2011) and the hybrid estimation of distribution algorithm with cheap and expensive local search (EDA/LS) (Zhou, Sun, and Zhang 2015) to study its performance. |
| Researcher Affiliation | Academia | Aimin Zhou,1 Jinyuan Zhang,1 Jianyong Sun,2 Guixu Zhang1 1East China Normal University, 3663 North Zhongshan Road, Shanghai, 200062, China 2Xi an Jiaotong University, No.28, Xianning West Road, Xi an, Shaanxi, 710049, China {amzhou, gxzhang}@cs.ecnu.edu.cn, jyzhang@stu.ecnu.edu.cn, jy.sun@xjtu.edu.cn |
| Pseudocode | Yes | A general FCPS assisted EA is shown in Algorithm 1. Algorithm 1: FCPS-EA Framework |
| Open Source Code | No | The paper does not provide any specific links to source code repositories, nor does it explicitly state that the code for the described methodology is released or available in supplementary materials. |
| Open Datasets | Yes | The first 13 benchmark functions from the YLL test suite (Yao, Liu, and Lin 1999) are employed for empirical study. |
| Dataset Splits | No | The paper uses benchmark functions (YLL test suite) as the problems to be optimized, not as a dataset that is split into training, validation, and test sets for the overall evaluation of the evolutionary algorithms. While the internal fuzzy classifier uses strategies like 'Mean fitness separation' to define training sets from the current population, this is an internal mechanism and not a global dataset split for reproducing the main experimental results. |
| Hardware Specification | No | The paper does not provide specific details regarding the hardware (e.g., CPU, GPU models, memory) used for conducting the experiments. |
| Software Dependencies | No | The paper mentions specific algorithms like Co DE and EDA/LS and methods like fuzzy KNN, but it does not list any specific software dependencies or library versions (e.g., Python 3.x, TensorFlow 2.x, PyTorch 1.x) that were used for implementation. |
| Experiment Setup | Yes | The variable dimensions are n = 30 for all instances. The population size is N = 150 for EDA/LS and its variants, and N = 30 for Co DE and its variants. The stop condition is fitness evaluations (FES) = 300,000 for all of the algorithms. Each algorithm is executed on each test instance for 30 independent runs. For FCPS based approaches, the number of candidate solutions is M = 4. The other control parameters are the same as in the original algorithms (Zhou, Sun, and Zhang 2015; Wang, Cai, and Zhang 2011). In the experiments, FCPS is also compared with BCPS... The MS strategy is used for training set definition. |