Topic: Data Mining on Streams

With the vast development of Internet Techonology, there is large amount of data transmitting online with very fast speed. Traditional data mining algorithms can no long satisfy our new efficiency and accuracy requirements. Since the data generating speed is much faster than the traditional mining speed, we strongly feel that we have no enough time to deal with all these data, also our unbounded memory requirements can no long fit in the limited memory space. All such constraints stimulate us to find out more appropriate algorithms to mininig on streams with constant memory and within constant time. And we have to require all data can be seen at most once. Here algorithm efficiency is very important for us. All coming data may represent a very complex model. If we only use samples, we may only get very simple model, which cannt represent the real currently accurate model. 

And for time-changing data, concept may drift with the time being. This requires our streaming model can detect all such changes, and refine the model simultaneously. 

Reading List:

Traditional Learning Algorithms

• S. M. Weiss, C. Apte, F. J. Damerau, D. E. Johnson, F. J. Oles, T. Goetz, T. Hampp. Maximizing Text-Mining Performance. 

• P.E. Utgoff. Incremental Indection of Decision Trees. Machine Learning, 4:161-186, 1989
• Avrim Blum. On-Line Algorithms in Machine Learning[survey]. CMU

• B. Babcock, S. Babu, M. Datar, R. Motwani, J. Widom. Models and Issues in Data Stream Systems[survey]. 
• P. Domingos, G. Hulten. Mining High-Speed Data Streams. Proceedings of the Sixth International Conference on Knowledge Discovery and Data Mining (pp. 71-80), 2000. Boston, MA: ACM Press.
• P. Domingos, G. Hulten, L. Spencer. Mining Time-Changing Data Streams. Proceedings of the Seventh International Conference on Knowledge Discovery and Data Mining (pp. 97-106), 2001. San Francisco, CA: ACM Press.
• P. Domingos, G. Hulten. A General Method for Scaling Up Machine Learning Algorithms and its Application to Clustering. Proceedings of the Eighteenth International Conference on Machine Learning (pp. 106-113), 2001. Williamstown, MA: Morgan Kaufmann.
• P. Domingos, G. Hulten. Catching Up with the Data: Research Issues in Mining Data Streams. Workshop on Research Issues in Data Mining and Knowledge Discovery, 2001.
• S. Guha, N. Mishra, R. Motwani, L. O'Callaghan. Clustering Data Streams. IEEE Symposium on Foundations of Computer Science, 2000.
• A. C. Gilbert, Y. Kotidis, S. M. Muthukrishnan, M. J. Strauss. Surfing Wavelets on Streams: One-Pass Summaries for Approximate Aggregate Queries. The VLDB Journal, 2001.
• Jon Kleinberg. Bursty and Hierarchical Structure in Streams[Concept Drift]. Cornell. Proceedings of the 8th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2002.