Dr. Mhd. Reza M. I. Pulungan, a lecturer at the UGM Faculty of Mathematics and Natural Sciences (FMIPA), was inaugurated as a professor of computer science on Tuesday (20/6).
Professor Pulungan delivered a speech entitled “Challenges in Modeling and Analysis of Continuous-Time Markov Chains and Their Solutions.” He stated that probabilistic verification is a part of formal methods in computer science.
In this field, the focus is not only on examining the functional correctness of various systems, including hardware, software, control systems, and safety-critical computerized systems, but also on studying stochastic behavior and real-time aspects.
“In our daily lives, we increasingly rely on such systems, so the certainty of their correctness is essential. In many analyses, the stochastic behavior and real-time aspects of such systems are studied using continuous-time Markov chains,” said the newly appointed professor.
Professor Pulungan mentioned that a Markov chain is a stochastic process, a sequence of random variables representing the changing configuration of a system over time.
However, continuous-time Markov chains are widespread in various fields of study because their modeling is natural and intuitive, allowing the essence of real natural phenomena to be captured by straightforward abstractions.
He emphasized the need for robust, stable, and accurate transient analysis methods because the discovered Markov models in real systems can be huge, with hundreds of thousands to millions of states.
The professor stated that he and his colleagues are implementing an algorithm developed into an online tool called APHMIN. However, this developed algorithm only guarantees that it almost always produces a minimal continuous-time Markov chain.
Therefore, they are currently working on improving the stability and speed of the algorithm by considering the use of embedded discrete-time Markov chains, which have stochastic matrices and eliminate the qualification of “almost always.”
Author: Gusti Grehenson
Photo: Firsto