Multi-Label Causal Feature Selection
Authors: Xingyu Wu, Bingbing Jiang, Kui Yu, Huanhuan Chen, Chunyan Miao6430-6437
AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on real-world data sets validate that MB-MCF could automatically determine the number of selected features and simultaneously achieve the best performance compared with state-of-the-art methods. |
| Researcher Affiliation | Academia | 1School of Computer Science and Technology, University of Science and Technology of China. 2Hangzhou Institute of Service Engineering, Hangzhou Normal University. 3School of Computer and Information, Hefei University of Technology. 4School of Computer Science and Engineering, Nanyang Technological University. |
| Pseudocode | Yes | Algorithm 1 The MB-MCF Algorithm. |
| Open Source Code | No | The paper does not provide any specific links or explicit statements regarding the public availability of its own source code for the methodology described. |
| Open Datasets | Yes | Table 1: Details of the multi-label data sets. Data set domain #Training #Test #Features #Labels cardinality density Birds audio 500 100 260 19 1.014 0.053 CAL500 music 300 100 68 174 26.044 0.150 EUR-Lex text 5000 2000 5000 201 2.213 0.011 Mediamill video 1000 1000 120 101 4.376 0.043 NUS-WIDE images 10000 5000 500 81 1.869 0.023 |
| Dataset Splits | Yes | Table 1: Details of the multi-label data sets. Data set domain #Training #Test #Features #Labels cardinality density Birds audio 500 100 260 19 1.014 0.053 CAL500 music 300 100 68 174 26.044 0.150 EUR-Lex text 5000 2000 5000 201 2.213 0.011 Mediamill video 1000 1000 120 101 4.376 0.043 NUS-WIDE images 10000 5000 500 81 1.869 0.023 |
| Hardware Specification | No | The paper does not provide specific details about the hardware (e.g., GPU/CPU models, memory) used for running its experiments. |
| Software Dependencies | No | The paper mentions using Hiton-MB as the MB discovery algorithm, G2-test for conditional independence tests, and ML-k NN for classification, but does not provide specific version numbers for any of these software components. |
| Experiment Setup | Yes | In addition, we also use the original data with no feature selection as a baseline in each experiment. To evaluate the effectiveness of the proposed methods, we employ a representative multi-label classification algorithm, ML-k NN (Zhang and Zhou 2007), to compute the classification accuracies archived by using selected features, and the number of nearest neighbors k is set to 10 with default setting. For a fair comparison, the regularization parameters for all comparing algorithms are tuned from t0.01, 0.1, 0.3, . . . , 0.9, 1u by grid search. |