Semi-supervised Feature Selection via Rescaled Linear Regression
Authors: Xiaojun Chen, Guowen Yuan, Feiping Nie, Joshua Zhexue Huang
IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results on eight real-life data sets show the superiority of the method. |
| Researcher Affiliation | Academia | 1College of Computer Science and Software, Shenzhen University, Shenzhen 518060, P.R. China 2School of Computer Science and Center for OPTical IMagery Analysis and Learning (OPTIMAL), Northwestern Polytechnical University, Xi an 710072, P. R. China |
| Pseudocode | Yes | Algorithm 1 Algorithm to solve problem (12); Algorithm 2 Algorithm to solve problem (1): RLSR |
| Open Source Code | No | The paper does not provide an explicit statement or link for open-source code availability. |
| Open Datasets | Yes | The 8 benchmark data sets were selected from Feiping Nie s page1. The characteristics of these 8 data sets are summarized in Table 1. 1http://www.escience.cn/system/file?fileId= 82035 |
| Dataset Splits | Yes | For the selected features, we first performed 10-fold cross-validation to select the best SVM model, then we tested the selected SVM model on the test part. For each of the 8 data sets, the training examples were randomly selected with the given ratio {10%, 20%, 30%, 40%, 50%}. The remaining examples were then used as the test data. |
| Hardware Specification | No | The paper does not provide any specific hardware details such as GPU/CPU models or other processor specifications used for running experiments. |
| Software Dependencies | No | The paper mentions using SVM but does not specify any software libraries or their version numbers used in the experiments. |
| Experiment Setup | Yes | We set the regularization parameter γ of LS, LSDF, RFS, UDFS, s Select and RLSR as {10 3, 10 2, 10 1, 1, 102, 103}, λ of s Select as {0.1, 0.2, 0.3, 0.4, 0.5, 0.6}. The projection dimensions for LS, LSDF, s Select and UDFS were set empirically around d/3 in our experiments, where d is the number of features in the data. |