DAMSS 2016 was December 1-3, 2016. The challenge of organizing a scientific event on such focused subjects was again completely achieved: DAMSS 2016 had 9 invited expertise lecturers, 15 oral presentations distributed by 5 sessions, 54 poster presentations in a total of 135 participants from 11 countries representing academia, research institutes.
Participants of DAMSS 2016
DOI: http://dx.doi.org/10.15388/DAMSS.2016
DAMSS 2016 : Invited Speakers
 |
Professor of statistics in the school of mathematics at Cardiff University. He is the author of 9 monographs, 11 edited volumes, more than 100 papers in refereed journals and about 100 papers in edited volumes and conference proceedings. His research interests include time series analysis, stochastic global optimisation, probabilistic methods in search and number theory.
|
Talk title: Some geometrical and analytical features of problems involving big data and high dimensions
Abstract: In this talk, I will concentrate on the following aspects of the science of big data science and large dimensions. First, I will outline some peculiar features of high-dimensional spaces. This will include a discussion on some counter-intuitive but very practically useful properties of the multivariate uniform and normal distributions. Second, I will discuss the so-called 'curse of dimensionality' in Monte-Carlo methods and global random search. I will try to demonstrate that it is impossible to guarantee any acceptable accuracy for high-dimensional black-box global optimization problems in the absence of Lipschitz-type information about the objective function. Third, I will discuss the problem of finding the most influential predictors in linear regression with a very large number of explanatory variables. Finally, I will discuss some big-data practical problems; the main attention will be paid to the problem of adaptive targeting in online advertising and effective methods of feature selection.
 |
Vice-Rector for Research and Transfer,Professor of Computer Science with the Department of Computer Science and Artificial Intelligence, University of Granada.
Current research interests include group decision making, consensus models, linguistic modeling, aggregation of information, information retrieval, bibliometric, digital libraries, web quality evaluation, recommender systems, and social media. He is the author of more than 170 papers in refereed journals and about 400 papers in edited volumes and conference proceedings.Around 16 of his papers are classed as highly cited in the Thomson Reuters database as well as being in the top 1% of the most cited papers in its field (Computer Science and Engineering). |
Talk title: Intelligent Decision Making and Consensus
Abstract: We analyze the fuzzy decision making models as to develop intelligent decision making processes in the real world. In particular, we focus on the soft consensus models by analyzing an overview of consensus models based on soft consensus measures, showing the pioneering and prominent papers, the main existing approaches and the new trends and challenges.
 |
J. Calvin is a professor in the department of computer science at the New Jersey Institute of Technology. He received PhD in Operations Research from Stanford University, an MS in the same field from the University of California at Berkeley, and an AB in Mathematics from Berkeley. His research interests are in global optimization and probabilistic analysis of algorithms.
|
Talk title: The Bayesian Approach to Global Optimization
Abstract: Many problems in engineering and science require finding parameter values that minimize a cost function. If no information is available about the number and location of minimizers (in particular if the function is not known to be convex) then the problem is a global optimization problem. For such problems, the information available to the optimizer, such as function values and derivatives at a finite number of parameters values, is insufficient to know the optimizing values and one must settle for an approximation. A popular approach to approximating solutions of global optimization problems is to adopt a probability model for the unknown cost function. The probability model motivates optimization algorithms, and also allows for analyzing the average approximation error of algorithms. In this talk, we describe some of the algorithms that have been proposed based on probability models, and some average-case complexity results.
 |
Panos is a Manager in EY’s Performance Improvement Advisory practice. He is a member of the CSE (Central and South-East Europe) Data Analytics and Enterprise Intelligence Center of Excellence, based in Athens. He has expertise in price, promotion and media modelling & analysis, data analysis, and has extensive experience in international consumer goods markets. Panos holds a PhD and MSc in Applied Mathematics and Complex Systems, an executive MBA, and a BSc in Pure & Applied Mathematics. He has worked with IRI in one of the biggest Center of Analytics internationally. In the last 3 years he was leading the global delivery team of IRI. His team was delivering marketing mix and forecasting projects in most of the major international consumer goods markets.
|
Talk title: Commercial Analytics to Improve Business Performance through Micro Customer Segmentation, Consumer Behaviour Analysis and Advanced Analytical Methods Application.
Abstract: Market dynamics are pushing companies across different industries to re-evaluate their Pricing (Budget) and Marketing Strategies in order to improve performance. Revenue and Profit growth objectives are now tighten with tougher margins, cost reduction and the necessity of higher value creation. Analytical services propose innovative solutions to its recipients and offer not only financial numbers growth but also strong customer relationships through enhanced customer engagement. Revenue Management and advanced analytics solutions are now in the front line for creating actionable value for our clients. The reason behind this is that analytics brings customer centricity and better decision-making to the business. So companies shift data resources to analytics for self-service Analytics-On-Demand, streamlined decision engine, prioritize delivery over data. The most important targeted (and specific) service that analytics can offer is the use of BI reporting tools that will help management keep track of essential metrics, the statistical analysis that will reveal causal relationships, the implementation of efficient and effective data management across the whole data lifecycle and above all how to embed a more holistic analytics strategy within each organization. Commercial analytics refer to a range of solutions that is mainly applied in the consumer packaged goods and Retail industry, but its techniques and learnings are applicable to many other industries, such as pharmaceutical, automotive, telco’s even banking. Analytics leverages past data to evaluate the underlying relationship between data input and output. The goal is to understand what drives customer and consumer behavior, what affects shopping patterns, and most importantly quantify the impact on sales, revenue and profit of various pricing, marketing but also competitive activities. During the presentation we will exhibit real world case of how applications of analytics helped our clients to get actionable insights, design and implement new strategies. In order to effectively do this and drive better decisions we must first ask the right business questions and then seek answers in the data.
 |
Florin Gheorghe FILIP took his MSc and PhD in control engineering from the TU “Politehnica” of Bucharest. In 1991 he was elected as a member of the Romanian Academy (RA). He has been a scientific researcher at the National R&D Institute in Informatics (ICI) of Bucharest. Currently, he is a part-time researcher at the National Institute of Economic Researches (INCE) of the RA, also the director of the Library of the Academy. He was elected as vice-president of RA in 2000 and reelected in 2002 and 2006. His main scientific interests include large–scale systems, decision support systems, technology management and foresight.
|
Talk title: Modern Information and Communication Technologies and their Impact on Decision Support Systems
Abstract: In recent years, important progress has been noticed in information and communication technologies (I&CT). The new technologies have had seriously influenced the industry business models and human skills and knowledge. Collaboration engineering, a new methodology, got traction in the new context. The paper aims at reviewing such developments with a particular emphasis on the collaborative decision-making activities in control and management (C&M) settings. The attributes of the current day organization, which is characterized by an ever increased degree of intra- and inter-enterprise collaboration, are reviewed first and the transition of the C&M schemes from genuine multilevel hierarchical structures to ever more cooperation schemes are described. Several advanced key technologies, such as BI&A (business intelligence and analytics), web technology, social networks, mobile and cloud computing that enable collaboration are reviewed and their impact on collaborative management and control activities are discussed.
 |
From 1991 he is affiliated with the Systems Research Institute of the Polish Academy of Sciences, from which he received his scientific degrees. He is also Professor in Graduate School for Applied Informatics and Management in Warsaw. The scientific interests are in optimization, multiple criteria decision making, computer-aided decision making, and also in identification, quantification, and management of risk in business organizations. Professor published over 100 scientific research papers, two research monographs and over 30 popular papers on risk management in finance and electrical energy trading.
|
Talk title: The Strength and Beauty of Optimization: the Case of Radiotherapy Planning in the Service of the Society
Abstract: Radiotherapy planning is a key stage in a chain of actions preceding any cancer radiation therapy and it gives rise to large-scale multiobjective problems. The battle here is for manageable problem sizes, solution accuracy and planners' agility in selecting clinically acceptable solutions (irradiation plans). As cancer diseases put nowadays a heavy toll on societies worldwide, the outcome of this battle is of importance to be measured on societal scales.
In the presentation, we shall make an attempt to list issues which should be addressed from the large-scale multiobjective perspective to improve the process of radiotherapy planning. It is expected that the tangible gains of the efforts will be twofold: better radiation therapy treatment effects, but also more effective usage of equipment, personnel, and resources of radiation oncology departments.
We also present our personal experience from a cooperation with an oncology clinic aimed at radiotherapy planning enhancements.
 |
General Director of the United Institute of Informatics Problems of the National Academy of Sciences of Belarus, Belarus State Prize Laureate (2002), senior IEEE member and IAPR member, Belarus representative in IAPR. His research subjects include image processing and analysis, medical imaging, mathematical morphology, bioinformatics, remote sensing, discrete applied mathematics. He is an author of more than 200 publications in the field of image processing, mathematical morphology, bioinformatics, discrete optimization.
|
Talk title: Computer Screening and Modeling for Anti HIV-1 Drug Development
Abstract: Developing new drugs against HIV is an area of intensive research. From more than 30 drugs approved today for use in clinical medicine for the treatment of HIV-1, only two could prevent the penetration of the virus into the target cell. The rest can act on the later stages of the virus replication cycle when it has already penetrated into the cell. About 10 years ago there were discovered antibodies that can neutralize various strains of HIV-1. The antibodies VRC01 and 10E8, for example, can neutralize more than 90% of viral isolates. The report is devoted to the presentation of computer technology that allows to find small chemical compounds that can interact with HIV-1 envelope proteins similarly to broad neutralizing antibodies (3074, VRC01, 10E8). The technology contains database search for finding potential chemical compounds, molecular docking and dynamics for evaluation of their inhibitory activity. It takes a lot of computational resources to implement the above simulation stages. The choice of the antibodies was not accidental. The antibody VRC01 blocks the first stage of HIV entry into a target cell, i.e. the virus binding to the primary CD4 receptor of the target cells. The antibody 3074 blocks the second step – the binding to co- receptors CCR5 or CXCR4. Finally, antibody 10E8 targets virus envelope protein gp41, which is responsible for the fusion of the virus membrane with the host cell membrane. We propose 18 compounds that can be considered as promising basic structures for the rational design of novel, potent, and broad-spectrum anti-HIV-1 drugs.
 |
Associate Professor of Almeria University. Her interest has been focused on high-performance computing, GPU computing, sparse matrix computation. FastSparse, tomography and image processing, scientific computation. She published 29 scientific research papers in International journals with Impact Factor (ISI).
|
Talk title: Energy-Aware Scientific Computing
Abstract: Nowadays an intensive effort is being developed to design approaches for improving energy/power efficiency of computational devices and platforms. Energy costs represent a relevant share of the total costs of High-Performance Computing (HPC) systems. They include several kinds of processing units, such as CPU cores and GPU, whose energy consumption depends on the kind of processing which is being performed. The tutorial revises the main resources to measure power/energy on architectures and the approaches to improve the energy efficiency of HPC systems. Then, we focus on approaches based on resources selection, which optimize energy efficiency on multicore and/or multi-GPU architectures. Moreover, a tool is analyzed to automatically find the optimal resources on heterogeneous platforms when iterative algorithms are executed. The approach allows automatically adapting the resources selection to the combination platform-resources/problem-size. This way the energy efficiency is optimized without previous knowledge about the HPC system.
 |
Professor of University of Turku, Finland, leading an Interaction Design Group at the Department of Information Technology. Prof. Sutinen's research interests are in the uses of information and communication technologies for development (ICT4D), computer science and ICT education, creative problem-solving, contextual design, and text tools for learning. He has been designing innovative educational approaches in developing countries and for special education. He has a 6 M Euros portfolio of R&D projects, funded by various EU sources, Academy of Finland and the Finnish Funding Agency for Technology and Innovation Tekes. He has supervised or co-supervised 19 PhDs.
|
Talk title: Digital Theology: A Computer Scientist's View
Abstract: Brief seminar on Digital Theology. It will be a good opportunity to hear about one of the more unusual interdisciplinary research areas involving advanced computing and digital technologies. Professor Sutinen's talk will discuss the use and development of digital tools such as human language technologies and analytics for solving theological problems and inter-faith dialogue to serve as instruments of peace.
 |
Full professor of technology enhanced learning in the Department of Educational Research, Lancaster University, UK. He is the Director of the Centre for Technology Enhanced Learning and the Director of Studies for the Doctoral Programme in e-Research and Technology Enhanced Learning, which recruits some 25 doctoral students annually, and is currently supporting some 100 students worldwide. Don’s main concern is with learning, and how digital technologies can support learning and teaching. He has conducted over 60 studies in the past 12 years, identifying innovative as well as successful and effective practices, in classrooms, after-school activities, and home and community settings. Commissioned studies have informed policy and practice, for EU and UK government departments and agencies, national support agencies, regional and local authorities, corporations including the BBC, and a wide variety of companies. He has published widely; within his total output of currently 183 publications, he has singled authored and co-authored 14 peer-reviewed journal articles, 11 monographs, books or special issues, 46 book chapters, 77 reports to funders, and 14 articles in professional journals. He is a long-standing member of the International Federation for Information Processing, is vice-chair of their Technical Committee on Education, vice-chair of their Working Group on Information Technology in Educational Management, and in 2014 received an Outstanding Services Award for his contribution to the field.
|
Talk title: Learning, Data Analysis and Software Systems – Qualitative and Quantitative Dilemmas?
Abstract: : Educational software is often now run through linked or integrated digital management systems that collect increasing quantities and forms of background data. Frequently as a minimum, when users log on, where they log on from when they select pages to view or read, how often they stay on a page when they click on links, who else they contact about their learning or work, and when they log off, are all collected. Arguments are made that these forms of data can be used to interpret features of users that will inform better learning. Some previous research studies have focused on learning features of the individual, using background data in qualitative ways (sometimes displaying outcomes through forms of imagery). Other previous studies have focused on much wider sets of data, gathered from across (sometimes very large) numbers of users; interpretations of those data are sometimes stated to say something about an individual’s learning from a statistical or quantitative perspective. This paper explores these different paradigms and perspectives, and through related dimensions concerned with making choices when analysing learning from data in software systems, argues that dilemmas in choice of methodological approach are not the only or necessarily the key dilemmas that researchers face if their research findings are to be of value to the field of education and learning. Different stakeholders – policy makers, educational advisers, head teachers or principals, teachers, parents, students, educational software developers – all need specific forms of data output if they are individually to be most effectively supported in terms of enhancing learning or teaching. I will use a number of studies and their outcomes to illustrate a current gap in our research concern, and consider future implications and dilemmas we face in this field.