Markov Chain Analyzer
Tool for modeling, analytical evaluation, and simulation of discrete and continuous Markov chains
A web application for designing and evaluating discrete and continuous-time Markov chains. Used in systems simulation courses, it enables graph-based modeling, steady-state probability computation, absorbing state analysis, and transient behavior simulation.
State transition graph editor and steady-state probability calculation
Project Overview & Context
The application combines a visual graph editor with numerical and analytical solvers. Users construct models interactively or import transition matrices. The tool verifies structural properties (ergodicity, reducibility, periodicity), solves for steady-state probabilities via linear systems, and plots state distribution trajectories over time.
Key Features & Capabilities
Transition Graph Editor
Visual editing of states and transitions with probability or transition rate parameterization.
Discrete & Continuous Time
Support for discrete-time Markov chains (DTMC) and continuous-time processes (CTMC with generator matrix Q).
Steady-State Distribution
Analytical calculation of limiting probabilities by solving the linear system pi * P = pi.
Absorbing Chain Analysis
Computation of fundamental matrices, mean time to absorption, and absorption probabilities.
LaTeX & Report Export
Exports matrices and equation systems to LaTeX syntax, with report export to HTML and PDF.
Educational Value & Pedagogical Context
Solving steady-state systems by hand for chains with more than four states is tedious and error-prone. The application allows students to check their solutions, observe how parameter variations affect stability, and compare theoretical limits with stochastic simulation runs.
Supervised Theses & Student Authors
Developed as a Master's thesis defended at the Department of Computer Science and Engineering, UWB, supervised by Richard Lipka:
Web tool for creating, editing, visualising and analysing Markov models
The goal of this thesis is to create a web tool for work with Markov chains with discrete and continuous time. The tool is written in Typescript and uses the Cytoscape graph library for model visualization and interaction. The tool allows an easy creation of chains, both by adding states and transitions sequentially, and by using a descriptive language. It is possible to perform a simulation over the created chains, which is visualized by animation and its progress can be observed.
