Evolutionary Classifier Chain for Multi-Dimensional Classification
Authors: Yu-Yang Zhang, Bin-Bin Jia, Min-Ling Zhang
AAAI 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Comparative experiments on 14 real datasets validate that ECCO outperforms 7 state-of-the-art MDC approaches. In this section, we conduct comparative experiments with 7 comparison algorithms on 14 real datasets. |
| Researcher Affiliation | Academia | 1School of Computer Science and Engineering, Southeast University, Nanjing 210096, China 2College of Electrical and Information Engineering, Lanzhou University of Technology, Lanzhou 730050, China 3Key Laboratory of Computer Network and Information Integration (Southeast University), Ministry of Education, China EMAIL, EMAIL, EMAIL |
| Pseudocode | Yes | Algorithm 1: Genetic Operators in ECCO Algorithm 2: Overall Framework of ECCO |
| Open Source Code | Yes | 1Code package and supplementary material are publicly available at: http://palm.seu.edu.cn/zhangml. |
| Open Datasets | Yes | In order to fully validate the effectiveness of the proposed algorithm, we selected 14 real datasets to examine the classification performance of our algorithm on the MDC problem. The number of examples (#Examples), the number of class dimensions (#Dim.), the number of labels in each class space (#Labels/Dim.) and the number of features (#Features) for each dataset are shown in Table 1. |
| Dataset Splits | Yes | Each dataset is partitioned into the training set and the test set in the ratio of 80% and 20%. It is worth noting that this division is kept constant for different comparison algorithms on the same dataset. This ensures the fairness of the experiments. In the training phase, 5-fold cross-validation is performed to avoid bias in the evaluation of solutions on the validation set (Liang et al. 2024b). |
| Hardware Specification | Yes | All experiments are executed with 50 parallel cores invoked on the same workstation with two Intel(R) Xeon(R) Gold 6230 CPUs. |
| Software Dependencies | No | The paper mentions using a Support Vector Machine (SVM) as the base classifier for each comparison algorithm, but does not specify the version of SVM or any other software libraries used. It also refers to GA operators without specifying their software implementation or version. |
| Experiment Setup | Yes | For each of the seven compared algorithms, the parameters in the code are consistent with the version in the corresponding paper respectively. We use a Support Vector Machine (SVM) as the base classifier for each comparison algorithm. For the proposed algorithm ECCO, the population size is empirically set to 200. And the number of iterations is set to 100 to balance diversity and computational efficiency (Liang et al. 2024a). |