•  
  •  
Collect from 网页模板

Speakers

 

Keynote Speaker

 

Prof. Sunil Vadera
Computer Science at University of Salford, UK

 

Sunil Vadera is a Professor of Computer Science and the Head of the School of Computing, Science and Engineering at the University of Salford. He is a Fellow of the British Computer Society, a Chartered Engineer (C.Eng) and Chartered IT Professional (CITP). He gained a first class BSc(Hons) in Computer Science and Mathematics from the University of Salford in 1982, receiving three best student prizes. Following graduation, he began his career as a Research Assistant and progressed to a Lectureship in Computer Science in 1984. He holds a PhD from the University of Manchester in the area of Formal Methods of Software Development which was awarded in 1992. He was promoted to a Senior Lecturer in 1997 and to a Chair in Computer Science in 2000. 

Sunil was chair of the British Computer Society Academic Accreditations Committee with responsibility for professional accreditation of all UK University programmes in Computer Science from January 2007 to December 2009. His research is driven by the desire to close the gap between theory and practice in Artificial Intelligence, with expertise in the areas of Bayesian networks, decision tree learning, credit assessment and planning.

 

Speech Title: From Bagging to Bandits for Cost-Sensitive Decision Tree Learning


Abstract: Decision tree learning is one of the major success stories of AI, with many data mining tools utilizing decision tree learning algorithms. Recent research in this field has been influenced by realizing that human decision making is not focused solely on accuracy, but also takes account of the potential implications of a decision. For example, a chemical engineer considers the risks of explosion when assessing the safety of a process plant, a bank manager carefully considers the implications of a customer defaulting on a loan and a medical consultant does not ignore the potential consequences of misdiagnosing a patient.

This realisation has led to significant interest in developing cost-sensitive decision tree learning algorithms. This key note presents a tour of the rich variety of cost-sensitive decision tree algorithms, aimed at illuminating the characteristics of the algorithms that will help researchers position their own work and identify gaps for future research. The key note will begin with early algorithms that make minor changes to the entropy based selection measure used in C4.5, present use of genetic algorithms to evolve cost-sensitive trees, describe the use of bagging and boosting, and conclude with recent work that explores ideas such as non-linear trees and multi-arm bandits. The presentation will be based on the authors work with colleagues and PhD students over the last decade, some of which is reported in the following publications:


• Sunil Vadera (2010), CSNL: A Cost-Sensitive Non-Linear Decision Tree Algorithm, ACM Transactions on Knowledge Discovery from Data, Vol 4, No 2, pp1-25.
• Lomax, S. and Vadera, S. (2011). An empirical comparison of cost-sensitive decision tree induction algorithms. Expert Systems, 28: 227–268
• Lomax, S. and Vadera, S. (2013). A survey of cost-sensitive decision tree induction algorithms, ACM Computing Surveys, Vol 45, No 2, pp1-35.
• Lomax, S. and Vadera, S. (2016) A Cost-Sensitive Decision Tree Learning Algorithm Based on a Multi-Armed Bandit Framework, The Computer Journal, DOI: https://doi.org/10.1093/comjnl/bxw015
• Nashnush, E. and Vadera, S. (2017) .Learning cost-sensitive Bayesian networks via direct and indirect methods, Integrated Computer-Aided Engineering, vol. 24, no. 1, pp. 17-26, 2017

Papers available from http://usir.salford.ac.uk/view/authors/13105.html
 

 

Plenary Speaker

 

Prof. Marat Akhmet
Middle East Technical University, Turkey
 

Marat Akhmet is a professor of mathematics at Middle East Technical University (Ankara, Turkey) known for his research on the chaos and bifurcation theory in differential equations and hybrid systems with applications in physics, neural networks, biology, medicine and economics . Born in Kazakhstan, he studied at Aktobe State University. He received his doctorate in 1984 at Kiev University . He has been awarded a Science Prize of TUBITAK (Turkey, 2015), for best achievments in scientific research. He is an author of four books: "Principles of Discontinuous Dinamical Systems", Springer, 2010, "Nonlinear Hybrid Continuous Discrete-Time Models", Atlantis Press (Springer), 2011, "Neural networks with Discontinuous Impact Activations," Springer, 2014, and "Replication of Chaos in Neural Networks, Economics and Physics", Springer&HEP, 2015. His has solved the Second Peskin conjecture for Integrate-and-fire biological oscillators, has introduced and developed theory of differential equations with piecewise constant argument of generalized type, many aspects of discontinous dynamical systems. The last decade his main subject of research is input-output analysis of chaos and irregular behavior of hybrid neural networks.
 

Speech Title: Poincare Chaos and Prospects for Neural Networks Research


Abstract: The presence of chaos is assumed as one of reasons for brain power and methods to increase functioning of robotics. Recently, we have introduced in papers [1-3] a new type of chaos. The phenomenon was called Poincare chaos. The chaos is initiated from new concepts of unpreictable point, unpredictable orbit and unpredictable function. By detailed discussion of the dynamics and application for differrential and discrete equations we will demonstrate how to extend neural networks research.


Reference:

1. M.U. Akhmet, M.O. Fen, Unpredictable points and chaos, Commun. Nonl. Sci. Num. Simul. 40 (2016) 1–5.
2. M.U. Akhmet, M.O. Fen, Existence of unpredictable solutions and chaos, Turkish Journal of Mathematics, accepted.
3. M.U. Akhmet, M.O. Fen, Poincare chaos and unpredictable functions, submitted.
4. Akhmet M. U., Fen M. O., Attraction of Li–Yorke Chaos by Retarded SICNNs (2015) Neurocomputing 147, pp. 330-42.
5. Akhmet M. U., Fen M. O., Attraction of Li–Yorke Chaos by Retarded SICNNs (2015) Neurocomputing 147, pp. 330-42.
6. Akhmet M., Fen M.O., Generation of cyclic/toroidal chaos by Hopfield neural networks (2014) Neurocomputing, 145, pp. 230-239.
7. Akhmet M. U., Fen M. O., Shunting inhibitory cellular neural networks with chaotic external inputs, Chaos, Volume 23, Issue 2, 2013, Article number 023112.
8. Akhmet M.U., Fen M.O., Period-Doubling Route to Chaos in Shunting Inhibitory Cellular Neural Networks, Proceedings of International Symposium on Health Informatics and Bioinformatics, 2013, 8th, September 25-27, Ankara, Turkey.
9. Akhmet M.U., Fen M.O. and Kivilcim A., Li-Yorke Chaos Generation by SICNNs with Chaotic/Almost Periodic Postsynaptic Currents (under second revision).
10. Akhmet M.U., Fen M.O., Entrainment by Chaos (2014) Journal of Nonlinear Science, 24 (3), pp. 411-439.
11. Akhmet M., Fen M.O., Chaotification of impulsive systems by perturbations (2014) International Journal of Bifurcation and Chaos, 24 (6), art. no. 1450078.
12. M.U. Akhmet, M.O. Fen, Replication of chaos in neural networks, physics and economy, Springer&HEP, 2015.