Sunday, November 15, 2015 (Tutorial I)
Mining Structured Sparsity Beyond Convexity
Pinghua Gong, Zheng Wang, Jiayu Zhou, and Jieping Ye
Sparse learning has become a very effective tool for mining useful patterns and information from data. It has been successfully applied to many real application problems and attracted a lot of attention from machine learning and data mining researchers in the last decade. The aim of this tutorial is to introduce both basics and recent advances in sparse learning, including (1) convex sparse learning models involving Lasso, group Lasso, sparse group Lasso, fused Lasso, graph fused Lasso, tree Lasso, etc. and modern optimization algorithms such as first-order/second-order batch-mode optimization algorithms and recently developed stochastic optimization algorithms, (2) non-convex sparse learning models such as non-convex regularized problems, matrix factorization, matrix completion, etc. and non-convex optimization algorithms involving DC programming, general iterative shrinkage and thresholding, proximal alternating linearized minimization and greedy algorithms, and (3) applying sparse learning to interesting real application problems in recommendation, healthcare analysis, etc.
Monday, November 16, 2015 (Tutorial II)
Scalable learning of graphical models
9:00AM - 12:30PM
François Petitjean and Geoffrey I. Webb
From understanding the structure of data, to classification and topic modeling, graphical models are core tools in machine learning and data mining. They combine probability and graph theories to form a compact representation of probability distributions. In the last decade, as data stores became larger and higher-dimensional, traditional algorithms for learning graphical models from data, with their lack of scalability, became less and less usable, thus directly decreasing the potential benefits of this core technology. To scale graphical modeling techniques to the size and dimensionality of most modern data stores, data science researchers and practitioners now have to meld the most recent advances in numerous specialized fields including graph theory, statistics, pattern mining and graphical modeling. This tutorial will cover the core building blocks that are necessary to build and use scalable graphical modeling technologies on large and high-dimensional data.
Tuesday, November 17, 2015 (Tutorial - III)
Behavioral Modeling in Social Networks: from Micro to Macro
Meng Jiang and Peng Cui
The development of social networks has enabled the collection of behavioral data of unprecedented size and complexity. Modern social platforms have realized that great scientific and marketing values are contained in the millions of billions of behavioral records. How can we model users’ behaviors in social networks? What are the concepts and principles in modeling the complex behaviors? Can we develop efficient models for accurate behavior prediction and detection in social applications such as recommender systems, personalized search and social marketing? In this tutorial, we answer these questions by uncovering the contextual dependency, cross-domain and cross-platform properties, synchronized and abnormal characteristics, and many other patterns of users’ behaviors. We introduce recent advances in modeling complex behaviors from the perspectives of individuals, groups and cascades (from micro to macro) in social networks. Finally we summarize the tutorial with discussions on open issues and challenges about behavior modeling in social networks.