Systems and Means of Informatics

2017, Volume 27, Issue 1, pp 60-72


  • A. K. Gorshenin
  • V. Yu. Kuzmin


It is known that mixed probability models allow high-performance data-flow analysis in the information systems. However, the resulting estimated parameters of the model can vary significantly depending on the actual realization of algorithms. To solve this problem, the heterogeneous computing service MSM Tools implementing a number of methods based on the analysis of mixed probability models has been created. This approach satisfies the requirement of necessary virtualization between the actual hardware of a researcher and that of a computing service; so, it eliminates the necessity to install any additional software, as the computing service requires only a browser to perform calculations.
Main architectural and interface solutions that were used in the development of MSM Tools are described. Functionality of the service is demonstrated by the example of a profiler data analysis.

[+] References (20)

[+] About this article