Dr Slava Vaisman

Lecturer

Mathematics
Faculty of Science
r.vaisman@uq.edu.au
+61 7 336 53264

Overview

Radislav (Slava) Vaisman is a faculty member in the School of Mathematics and Physics at the University of Queensland. Radislav earned his Ph.D. in Information System Engineering from the Technion, Israel Institute of Technology in 2014. Radislav’s research interests lie at the intersection of applied probability, statistics, and computer science. Such a multidisciplinary combination allows him to handle both theoretical and real-life problems, in the fields of machine learning, optimization, safety, and system reliability research, and more. He has published in top-ranking journals such as Statistics and Computing, INFORMS, Journal on Computing, Structural Safety, and IEEE Transactions on Reliability. The Stochastic Enumeration algorithm, which was introduced and analyzed by Radislav Vaisman, had led to the efficient solution of several problems that were out of reach of state of the art methods. In addition, he is an author of 3 books with three of the most prestigious publishers in the field, Wiley, Springer, and CRC Press. Radislav serves on the editorial board of the Stochastic Models journal.

Research Interests

  • Data science
  • Statistics and Machine Learning
  • Rare Event Simulation and Modelling
  • System Reliability
  • Evolutionary Computation
  • Advanced Monte Carlo Methods and Randomized Algorithms
  • Stochastic Optimization and Counting
  • Graphical Models
  • Markov Decision Processes and Planning under uncertainty

Research Impacts

Radislav Vaisman’s research interests lie at the intersection of applied probability and computer science where he has made key contributions to the theory and the practical usage of Sequential Monte Carlo methods. Specifically, his work led to the publication of a book by John Wiley & Sons: Fast Sequential Monte Carlo Methods for Counting and Optimization, which covers the state-of-the-art of modern simulation techniques for counting and optimization. In addition, his contribution to the field of System Reliability resulted in the book: Ternary Networks: Reliability and Monte Carlo, by Springer. In 2019, Radislav coauthored the book: Data Science and Machine Learning: Mathematical and Statistical Methods, which was published by CRC Press. Dr. Vaisman has published in top-ranking journals such as Statistics and Computing, INFORMS, Journal on Computing, Structural Safety, Networks, and IEEE Transactions on Reliability.

Radislav Vaisman's research in the field of Sequential Monte Carlo led to the development of the Stochastic Enumeration method for estimating the size of backtrack trees. The proposed method tackles this very general but difficult problem in computational sciences. Dr. Vaisman also developed a rigorous analysis of the Stochastic Enumeration procedure and showed that it results in significant variance reduction as compared to available alternatives. In addition, he applied the multilevel splitting ideas to many practical applications, such as optimization, counting, and network studies. Dr. Vaisman has produced insightful work in the field of systems reliability, both in theory and practice. In particular, he has developed Sequential Monte Carlo methods for estimating failure probability in highly reliable structures and new sampling plans for estimating network reliability based on a network’s structural invariants. This contribution has been recognized by top scientific journals in this field, namely Structural Safety and IEEE Transactions on Reliability.

Qualifications

  • Doctor of Philosophy, Israel Institute of Technology
  • Master of Science, Open Uni Israel
  • Bachelor of Science, Technion, Israel Institute of Technology

Publications

View all Publications

Supervision

  • Doctor Philosophy

  • Doctor Philosophy

  • Master Philosophy

View all Supervision

Available Projects

  • I am always looking for prospective Ph.D. students. If you wish to know more about available projects, feel free to send me an email with your CV and a few lines regarding your research background and interests.

    For details, please see: https://people.smp.uq.edu.au/RadislavVaisman/Research.html

  • The majority of complex systems and products that empower our daily activities are subject to degradation. This affects the system lifetime, the quality of the service, and the corresponding safety of usage. Thus, a development of reliability management and prognostic programs is of overwhelming importance. In this project, you will investigate methods for understanding and managing of degradation processes. Specifically, the broad objective of this project is to develop new mathematical techniques and fast computational algorithms for inference in complex statistical models by building on recent advances in Monte Carlo methods, stochastic optimisation, and rare-event sampling techniques.

  • Statistical inference is one of the most important tools used for scientific investigation. When dealing with data, the Bayesian paradigm is very appealing since it allows to incorporate prior knowledge into a proposed model, provides a well-structured inference method (conditional on the newly observed information), does not rely on asymptotic approximation, provides interpretable answers, and implements a straight-forward framework for model comparison and hypothesis testing. While these merits often come with high computational costs, a continuing progress in the available computing resources allowed Bayesian statistics to rise to greater eminence in many scientific fields such as natural science, econometrics, social science, and engineering. However, despite recent advances, many real-life inference problems are still beyond the reach for classical Bayesian methods. Specifically, for many practical models, the evaluation of the likelihood function, a critical component of the Bayesian analysis, is either intractable or computationally prohibitive. In this project, you will investigate a number of methods such as the Pseudo-Marginal, the Integrated Nested Laplace, the Bayesian Synthetic Likelihood, the Variational Bayes, and the Approximate Bayesian Computation.

View all Available Projects

Publications

Book

Book Chapter

  • Gertsbakh, Ilya B., Shpungin, Yoseph and Vaisman, Radislav (2018). Reliability of a network with heterogeneous components. Recent Advances in Multi-state Systems Reliability. (pp. 3-18) Cham, Switzerland: Springer. doi: 10.1007/978-3-319-63423-4_1

Journal Article

Conference Publication

  • Moreno, Gabriel A., Strichman, Ofer, Chaki, Sagar and Vaisman, Radislav (2017). Decision-making with cross-entropy for self-adaptation. 12th IEEE/ACM International Symposium on Software Engineering for Adaptive and Self-Managing Systems, SEAMS 2017, Buenos Aires, Argentina, 22 - 23 May 2017. Piscataway, NJ, United States: Institute of Electrical and Electronics Engineers. doi: 10.1109/SEAMS.2017.7

  • Gertsbakh, Ilya B. , Shpungin, Yoseph and Vaisman, Radislav (2016). D-spectra for networks with binary and ternary components. Second International Symposium on Stochastic Models in Reliability Engineering, Life Science and Operations Management (SMRLO’16), Beer Sheva, Israel, 15 -18 Febuary 2016. Piscataway, NJ, United States: Institute of Electrical and Electronics Engineers. doi: 10.1109/SMRLO.2016.44

  • Salomone, Robert, Vaisman, Radislav and Kroese, Dirk (2016). Estimating the number of vertices in convex polytopes. 4th Annual International Conference on Operations Research and Statistics (ORS 2016), 5th Annual Conference on Computational Mathematics, Computational Geometry & Statistics (CMCGS 2016), Singapore, Singapore, 18 - 19 January 2016. Singapore, Singapore: Global Science and Technology Forum. doi: 10.5176/2251-1938_ORS16.25

  • Shah, Rohan and Vaisman, Radislav (2016). New sampling plans for estimating residual connectedness reliability. 4th Annual International Conference on Operations Research and Statistics (ORS 2016), City of Singapore, Singapore, 18-19 January 2016. Singapore: Global Science and Technology Forum. doi: 10.5176/2251-1938_ORS16.18

  • Botev, Zdravko I., Vaisman, Slava, Rubinstein, Reuven Y. and L’Ecuyer, Pierre (2014). Reliability of stochastic flow networks with continuous link capacities. 2014 Winter Simulation Confernce, Savannah, GA, USA, 7-10 December 2014. Piscataway, NJ United States: Institute of Electrical and Electronics Engineers. doi: 10.1109/WSC.2014.7019919

Grants (Administered at UQ)

PhD and MPhil Supervision

Current Supervision

  • Doctor Philosophy — Principal Advisor

    Other advisors:

  • Doctor Philosophy — Associate Advisor

    Other advisors:

  • Master Philosophy — Associate Advisor

    Other advisors:

  • Doctor Philosophy — Associate Advisor

    Other advisors:

  • Doctor Philosophy — Associate Advisor

Completed Supervision

Possible Research Projects

Note for students: The possible research projects listed on this page may not be comprehensive or up to date. Always feel free to contact the staff for more information, and also with your own research ideas.

  • I am always looking for prospective Ph.D. students. If you wish to know more about available projects, feel free to send me an email with your CV and a few lines regarding your research background and interests.

    For details, please see: https://people.smp.uq.edu.au/RadislavVaisman/Research.html

  • The majority of complex systems and products that empower our daily activities are subject to degradation. This affects the system lifetime, the quality of the service, and the corresponding safety of usage. Thus, a development of reliability management and prognostic programs is of overwhelming importance. In this project, you will investigate methods for understanding and managing of degradation processes. Specifically, the broad objective of this project is to develop new mathematical techniques and fast computational algorithms for inference in complex statistical models by building on recent advances in Monte Carlo methods, stochastic optimisation, and rare-event sampling techniques.

  • Statistical inference is one of the most important tools used for scientific investigation. When dealing with data, the Bayesian paradigm is very appealing since it allows to incorporate prior knowledge into a proposed model, provides a well-structured inference method (conditional on the newly observed information), does not rely on asymptotic approximation, provides interpretable answers, and implements a straight-forward framework for model comparison and hypothesis testing. While these merits often come with high computational costs, a continuing progress in the available computing resources allowed Bayesian statistics to rise to greater eminence in many scientific fields such as natural science, econometrics, social science, and engineering. However, despite recent advances, many real-life inference problems are still beyond the reach for classical Bayesian methods. Specifically, for many practical models, the evaluation of the likelihood function, a critical component of the Bayesian analysis, is either intractable or computationally prohibitive. In this project, you will investigate a number of methods such as the Pseudo-Marginal, the Integrated Nested Laplace, the Bayesian Synthetic Likelihood, the Variational Bayes, and the Approximate Bayesian Computation.

  • A series of interesting projects in the field of advanced Monte Carlo methods is available. In this project, you can expect to encounter various problems in the domains of Bayesian inference, time-series analysis, and modern machine learning.

  • In this project you will investigate a series of advanced statistical inference methods with application to crop yield. The methods range from time-series analysis and forecasting to artificial deep neural networks.

  • Spatial micro-simulation aims to generate a synthetic population from an anonymous sample data at the individual level, which matches the observed population in a geographical zone for a given set of criteria in the most realistic manner. A good micro-simulation method will allow to create estimated populations at a range of spatial scales where data may be otherwise unavailable. This project focuses on exploring efficient algorithms for spatial micro-simulation.