报告题目: Multi-Graph Learning
时间：2016年11月22日上午10点30
地点：学院B403
报告人：Jia Wu（吴佳）
报告人单位：澳大利亚悉尼科技大学
报告摘要:
Learning and classifying objects have been commonly used for many applications, such as image retrieval, text classification, and spam detection etc. For learning purposes, objects are required to be represented as instances by using feature vector and class label to denote characteristics and categories of the objects, respectively. In reality, the above simplified feature and class representation is inadequate for certain applications, which involve objects with unstable characteristics or complex behaviors. For instance, an image can be represented as a bag with each region inside the image denoting an instance inside the bag. In order to tackle the above mentioned complications, multi-instance (MI) is emerged as a new classification tool with each object provided for learning (or classifying) being a bag of instances. For MI learning, existing methods require that training samples are provided and represented in vector space, which inherently prohibits them from being applied to complicated objects containing structure information. To this end, we advance graph classification to handle multi-graph learning for complicated objects, where each object is represented as a bag containing several graphs and the label is only available for each graph bag but not individual graphs.
Jia Wu（吴佳）：澳大利亚悉尼科技大学讲师，博士、Research Associate、IEEE会员。主要研究领域为数据挖掘、机器学习、人工智能，及其在商业、工业、生物信息学、医疗信息学等领域的应用。迄今，在国际学术期刊和会议上共发表论文60多篇, 包括IEEE Transactions on Knowledge and Data Engineering、IEEE Transactions on Cybernetics、ACM Transactions on Knowledge Discovery Data、Pattern Recognition、IJCAI、AAAI、ICDM、SDM、CIKM、DASFAA等。曾获得2014顶级国际数据挖掘会议International Conference on Data Mining的最佳论文提名奖。
现任IEEE Access (SCI: 1.27), Journal of Next Generation Information Technology, International Journal of Engineering and Industries副主编和Complexity Journal (SCI: 3.514)客座主编。担任国际顶级神经网络大会2016、2017 International Joint Conference on Neural Networks的专题分会主席 (Special Session Chair)、顶级人工智能国际会议International Joint Conference on Artificial Intelligence 2017的高级程序委员 (Senior Program Committee)，顶级国际学术会议的程序委员 (Program Committee), 包括IJCAI、AAAI、ICDM、SDM、CIKM、DASFAA、PAKDD、IJCNN等。并应邀为多家顶级国际学术期刊和会议担任评审委员会委员，国际学期刊包括IEEE TKDE、IEEE TSMC、IEEEPAMI、IEEETNNLS、ACM TKDD、PR等，国际学术会议包括IJCAI、AAAI、ICDM、SDM、PAKDD、IJCNN、DASFAA、CIKM等。
邀请人：杜博 教授
