Duality of Mining Problems
Chair Professor Tzung-Pei Hong
Department of Computer Science and Information Engineering
National University of Kaohsiung, Taiwan
Frequent-itemset mining and erasable-itemset mining are two commonly seen and practical techniques for finding different useful itemsets in data mining. Frequent-itemset mining is a significant pre-processing step in the search for association rules and is mainly conducted based on the frequencies of the itemsets in a transaction database. On the other hand, erasable-itemset mining is often applied to product production planning and identifies the itemsets that would not significantly impact production profits if removed. Although the two mining techniques seem to be different and independent, we have derived they are actually equivalent to each other. We formally prove that these two mining techniques possess the property of duality. We design methods to transform one of the two mining problems into the other and then solve it, and vice versa. The mining results with and without the transformation will be identical. We then extend the duality property of the two mining problems to multiple-threshold and weighted-item situations. With the duality property, we can easily design an algorithm from its corresponding one in the dual mining problem.
Tzung-Pei Hong received his Ph.D. degree in computer science and information engineering from National Chiao-Tung University in 1992. He served as the first director of the library and computer center, the Dean of Academic Affairs, and the Vice President in the National University of Kaohsiung, Taiwan. He is currently a chair and distinguished professor at the Department of Computer Science and Information Engineering in NUK, and a joint professor at the Department of Computer Science and Engineering, National Sun Yat-sen University, Taiwan.
He has published more than 600 research papers in international/national journals and conferences and has planned more than fifty information systems. He is also the board member of more than forty journals and the program committee member of more than one thousand conferences. His current research interests include knowledge engineering, data mining, soft computing, management information systems, and www applications.
Neverending Machine Learning
Professor Michał Woźniak
Department of Systems and Computer Networks, Faculty of Computer Science and Telecommunications
Wroclaw University of Science and Technology
Lifelong Machine Learning (LLML) can overcome the limitations of statistical learning algorithms that need many training examples and are suitable for isolated single-task learning. Key features that need to be developed within such systems to benefit from prior learned knowledge include feature modeling, knowledge retaining from past learning tasks, knowledge transfer to future learning tasks, previous knowledge updates, and user feedback.
Also, the concept of task that appears in many formal definitions of lifelong ML models seems hard to define in many real-life setups because it is often difficult to distinguish when a particular task finishes and subsequent starts.
One of the main challenges is the stability and plasticity dilemma, where the learning systems have to trade-off between learning new information without forgetting the old one. It is visible in the catastrophic forgetting phenomenon, which is defined as a complete forgetting of previously learned information by a neural network exposed to the new information.
The talk will focus on the main approaches related to Lifelong Machine Learning and the open challenges in this domain.
Michał Woźniak is a professor of computer science at the Department of Systems and Computer Networks, Wroclaw University of Science and Technology, Poland. His research focuses on machine learning, compound classification methods, classifier ensembles, data stream mining, and imbalanced data processing. Prof. Wozniak has been involved in research projects related to the topics mentioned above and has been a consultant for several commercial projects for well-known Polish companies and public administration. He has published over 300 papers and three books. Prof. Wozniak was awarded numerous prestigious awards for his scientific achievements as IBM Smarter Planet Faculty Innovation Award (twice) or IEEE Outstanding Leadership Award, and several best paper awards from the prestigious conferences.