Keynote Talk 1. Big Data: Analysis and Procedures
Part 1. Fuad Aleskerov “Choice Procedures in Big Data Analysis”
Fuad Aleskerov, Head of Mathematics Department of National Research University “Higher School of Economics” (HSE), Head of International Laboratory of Decision Choice and Analysis of HSE (http://www.hse.ru/en/org/persons/140159)
Areas of expertise: Decision theory, Social and individual choice theory, multicriteria choice theory, Theory of Binary Relations, Theory of Political Processes, Discrete Mathematical Models, Banking.
A set of choice procedures is presented for an analysis of big data, specially, in the search problem. These procedures include different versions of a superposition of super-threshold choice rules, and several other rules, including Pareto-rule and uni-criterial rule. It is shown that the proposed procedures perform better than known procedures, e.g., Support Vector Machines. A comparison of different procedures on Microsoft Data is presented.
Part 2. Sergiy Butenko "Network-Based Analysis of Big Data"
Sergiy Butenko, Associate Professor and Donna and Jim Furber `64 Faculty Fellow in Industrial and Systems Engineering, Ph.D., University of Florida
Dr. Butenko's research concentrates mainly on global and discrete optimization and their applications. In particular, he is interested in theoretical and computational aspects of continuous global optimization approaches for solving discrete optimization problems on graphs. Applications of interest include network-based data mining, analysis of biological and social networks, wireless ad hoc and sensor networks, and energy.
Big data arising in various complex systems can be conveniently modeled using networks, in which the components of a complex system are described by nodes and pairwise interactions between different components are represented by edges. Network-based analysis allows to capture some global structural properties of the system and predict overall trends in its dynamics. This talk will focus a methodological framework for analyzing clusters in networks representing big data.
Keynote Talk 2.
Panos M. Pardalos “A new embedded feature selection method for high dimensional datasets”
Panos M. Pardalos, Distinguished Professor Center For Applied Optimization, Industrial and Systems Engineering, University of Florida, USA, http://www.ise.ufl.edu/pardalos/
National Research University Higher School of Economics,
Laboratory of Algorithms and Technologies for Network Analysis, Russia, http://www.hse.ru/en/org/persons/44226286
Reseach interests: Global Optimization and Applications, Design and Analysis of Computer Algorithms, Computational Neuroscience, Parallel Computing in Mathematical Programming, Optimization in Biomedical Engineering, Telecommunications, Supply Chain, E-commerce, and Financial Engineering, Information Theory and Control, Massive Datasets and Data Mining, Cooperative Systems, Scientist Computing, Software Design and Development.
An efficient algorithm based on alternate optimization techniques is proposed. Numerical experiments on several publicly available datasets show that our proposed method can obtain competitive or better performance compared with other embedded feature selection methods. Moreover, sPSVMs remove more than 98% features in many high dimensional datasets without compromising on generalization performance and also show consistency in the feature selection process. Additionally, sPSVMs can be viewed as inducing class-specific local sparsity instead of global sparsity like other embedded methods and thus offer the advantage of interpreting the selected features in the context of the classes.
High Dimensional datasets are currently prevalent in many practical applications. Classification and feature selection are common tasks performed on such datasets. In this talk, a new embedded feature selection method for high dimensional datasets is introduced by incorporating sparsity in Proximal Support Vector Machines (PSVMs). Our method called Sparse Proximal Support Vector Machines (sPSVMs) learns a sparse representation of PSVMs by first casting it as an equivalent least squares problem and then introducing the l1-norm for sparsity.
Keynote Talk 3.
Shouyang Wang “TEI@I Methodology and Applications to Economic Analysis and Forecasting”
Center for Forecasting Science, Chinese Academy of Sciences;
Academy of Mathematics and Systems Science, Chinese Academy of Sciences; http://sourcedb.cas.cn/sourcedb_amss_cas/yw/rck/200907/t20090722_2134098.html
Management School, The University of Chinese Academy of Sciences
A methodology is introduced for study complex systems, especially economic and social systems. The methodology is named TEI@I Methodology which implies that four components are fundamental in studying complex systems. The methodology is illustrated via economic analysis and forecasting.