** This software package implements the algorithm proposed in the following paper:
       Zexi Huang,
       Transfer Learning for Community Detection.
       Bachelor thesis, University of Electronic Science and Technology of China, Chengdu, China, 2018.
   Please cite this paper if you find this software useful.

** Usage:
   Run transfer_learning_for_community_detection(outputFile,inputFile,targetLayerID,kCom,c,ck,T1,T2)
   where the parameters specifiy:
       outputFile: the file that contains the output community memberships in the target layer. Each line has the following format: nodeID communityID.
       inputFile: the file that contains the input connection information of the multiplex network. Each line shoule have the following format: layerID nodeID nodeID.
       targetLayerID: the layerID of the target layer that receives the transferred knowledge.
       kCom: number of communities to be detected in the targe layer.
       c: number of dimensions of common feature matrix.
       ck: number of dimensions of layer-specific feature matices.
       T1: maximum number of iterations of the learning algorithm.
       T2: maximum number of iterations of the clustering algorithm.

** A simple test:
   You can run
       transfer_learning_for_community_detection('seventhgrade.communities','seventhgrade.edges',3,3,3,[2 2 2],20,100)
   for transfer learning for community detection in the third layer of the Seventhgrade multiplex network (included in this package). The parameters used are the same to those in the referred paper. 

**	Please contact Zexi Huang (eitima@163.com) for further comments, suggestions and discussions.
