Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Maximum Margin Multi-Dimensional Classification
Authors: Bin-Bin Jia, Min-Ling Zhang4312-4319
AAAI 2020 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental studies over real-world MDC data sets clearly validate effectiveness of the proposed maximum margin MDC techniques. Experimental Results Table 3 reports the detailed experimental results of five comparing approaches in terms of each evaluation metric, where the best performance among all comparing approaches is shown in boldface. |
| 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 4Collaborative Innovation Center of Wireless Communications Technology, China EMAIL |
| Pseudocode | Yes | Table 1: The pseudo-code of M3MDC. |
| Open Source Code | No | The paper provides a link to LIBSVM, a third-party tool used, but does not state that the code for their proposed M3MDC method is open-source or provide a link to it. |
| Open Datasets | Yes | A total of ten benchmark data sets are employed for performance evaluation. Table 2 summarizes the characteristics of all MDC data sets, including number of examples (#Exam.), number of class spaces (#Dim.), number of class labels per class space (#Labels/Dim.),2 and number of features (#Features). (Datasets listed: Edm, Flare1, Cal500, Music, Song, WQplants, WQanimals, Water Quality, Yeast, Voice) |
| Dataset Splits | Yes | Ten-fold cross-validation is performed on the benchmark data sets, where the mean metric value as well as standard deviation are recorded for each comparing approach. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments. |
| Software Dependencies | No | The paper mentions 'MOSEK optimization software' and 'Matlab' and 'LIBSVM (Chang and Lin 2011)' but does not specify their version numbers. |
| Experiment Setup | Yes | As shown in Table 1, the two regularization parameters for M3MDC are set to be λ1 = 0.1, λ2 = 0.001 respectively. For ensemble approaches ECC, ECP and ESC, a random cut of 67% examples from the original MDC training set is used to generate the base MDC model and the number of base classifiers is set to be 10. Furthermore, predictions of base MDC models are combined via majority voting. Support vector machine (SVM) is used to instantiate BR, ECC, ECP, ESC as base classifier. Specifically, LIBSVM (Chang and Lin 2011) with linear kernel is used. |