IEEE ICDM 2013 Conference in Dallas, TX
The IEEE International Conference on Data Mining (ICDM) started on Sunday, Dec. 8 with a welcome by the conference co-chair: Prof. Dr. Bhavani Thuraisingham (see picture right). Even tough the conference had 500+ registrations, the number of attendees was much lower due to the winter weather with freezing temperatures. Following the welcome, the first keynote talk titled “Opportunities and Challenges facing Recommender Systems: Where can we go from here?” was held by Prof. Dr. Alexander Tuzhilin (NYU, see picture below).
Wednesday 11 December 2013
Finally, Petko Bogdanov (see picture below) presented their talk (co-authors: M. Monglovi & A. Singh) titled “Mining Evolving Network Processes”.
He discussed the state-of-the-art in recommender systems. He grouped them into first generation (1G), 2G, and context-award 3G recommender systems (CARS).
Next, I attended the Classification I session. The first paper was by A. Dukkipati, G. Pandey, D. Ghoshdastidar, P. Koley, and D.M.V.S. Sriram titled “Generative Maximum Entropy Learning for Multiclass Classification” (see pictures below).
Meng Fang, Yin Jie, and Xingquan Zhu presented their talk “Transfer Learning across Networks for Collective Classification” (see pictures below).
Next, Xu-Ying Liu, Qian-Qian Li & Zhi-Hua Zhou (see pictures right and below) presented their paper “Learning Imbalanced Multi-Class Data with Optimal Dichotomy Weights”.
Finally, F. Giannotti, L. Milli, A. Monreale, D. Pedreschi, G. Rossetti & F. Sebastiani presented their talk “Quantification Trees”.
After lunch, I attended the Graph and Network Mining session. The first talk I saw was titled “Blocking Simple and Complex Contagion By Edge Removal” (Ch. Kuhlman, G. Tuli, S. Swarup, M. Marathe, and S.S. Ravi).
Next, Danai Koutra (co-authors: H. Tong, D. Lubensky) presented their paper titled “BIG-ALIGN: Fast Bipartite Graph Alignment”.
During the breaks we had plenty of opportunities for networking.
That evening, the welcome reception was held (see pictures below). This also included a poster session.
Next, I attended the talk about “Statistical Selection of Congruent Subspaces for Outlier Detection on Attributed Graphs” (authors: P.I. Sanchez, E. Mueller, F. Laforet, F. Keller, K. Boehm).
On Monday, Dec. 9, the conference started with a keynote by Prof. Dr. Joydeep Ghosh (see picture below) titled “Predictive Healthcare Analytics under Privacy Constraints”. He covered several topics: availability of “Big” Data in healthcare, how to analyze such data, privacy-preserving measures.
Next, I joined the tutorial titled “Applied Matrix Analytics: Recent Advance and Case Studies” by H. Tong, F. Wang, Ch. Ding. They covered both theory (NMF: non-negative matrix factorization, CUR/CX decomposition, which features better interpretability than SVD, and the even better Colibri-S method) and applications.
Next, the focus shifted to the application of matrix math in health informatics. Dr. Fei Wang gave a very detailed talk on how to analyze Electronic Health Record data.
After lunch, all ICDM attendees boarded busses to either the Perot Museum of Nature and Science or the JFK tour to the Sixth Floor Museum (corner of Elm Str. and Houston). I chose the latter, so below you see some pictures of our tour.
That evening, the banquet was held at the Sheraton hotel. This also included an award ceremony, of course. Prof. Dr. Hans-Peter Kriegel (LMU Munich, Germany, see picture below) received the IEEE ICDM Research Contribution Award for outstanding research contributions, and Prof. Dr. Geoff Webb received the IEEE ICDM Outstanding Service Award. Congratulations!
On Tuesday, Dec. 10, Prof. Dr. JC. Mao (Microsoft, see picture below) presented his keynote talk titled “Large-Scale Learning in Computational Advertising”. The openness and detail of discussion is really appreciated! He discussed response prediction systems, multi-label random forests. In sum, great content!
After the coffee break, yours truly (@dirkvandenpoel) attended the “Social Media Mining: Fundamental Issues and Challenges” tutorial by Mohammad Ali Abbasi, Huan Liu, and Reza Zafarani (see pictures below).
During lunch, the ICDM community meeting was held.
Next, Prof. Dr. Shusaku Tsumoto (see picture below) took the stage to talk about the review process, which topics were accepted, ... .
Finally, I attended that ICDM Panel on Data Mining with Big Data. From left to right: Professors Geoff Webb (Monash University, Australia), Zhi-Hua Zhou (Nanjing University, China), Xindong Wu (University of Vermont), Chris Clifton (Purdue University assigned to the NSF), Bhavani Thuraisingham (University of Texas at Dallas), Vipin Kumar (University of Minnesota), and Jian Pei (Simon Fraser University, Canada).
