Readme for the MultiLabelCrowd Package
version July 2020
The package includes the code of four algorithms in [1], with two multi-label crowdsourcing classification algorithms NAM, RAM and two multi-label active crowdsourcing classification algorithms NAC, RAC.
[1] S.-Y. Li, Y. Jiang, N.V. Chawla, and Z.-H. Zhou. Multi-label learning from crowds. IEEE Transactions on Knowledge and Data Engineering 31(7), 1369{1382 (2019)
ATTN:
-
This package is free for academic usage. You can run it at your own risk. For other purposes, please contact Professor Zhi-Hua Zhou(zhouzh@nju.edu.cn).
-
This package was developed by Ms. Shao-Yuan Li (lisy@nuaa.edu.cn). For any problem concerning the code, please feel free to contact Ms. Li.
Code description: NAM.rar: The neirborhood aware multi-label crowdsourcing classifcation algorithm. To get demo results, run NAM/main.m in matlab.
NAC.rar: The active multi-label crowdosurcing learning algorithm extended from NAM. To get demo results, run NAC/main.m in matlab.
RAM.rar: The label relevance ware multi-label crowdsourcing classifcation algorithm. To get demo results, run RAM/main.m in matlab.
RAC.rar:The active multi-label crowdosurcing learning algorithm extended from RAM. To get demo results, run NAC/main.m in matlab.
data_sample.rar: one dataset sample used in [1], corresponds to the dataset1 in [1].
misc.rar: resource functions used by the four algorithms