Tarasyuk O. Methods and tools of metric-probabilistic quality assessment of the software for information and control systems of critical application

The dissertation is devoted to development of methods and tools of software quality and reliability assessment for the information and control systems of critical applications based on complex usage of the metrics and probabilistic models. Scientific results are: 1) a method of a complex assessment of software quality and reliability based on co-operative usage and integration of metrics and models outcomes, which allows to provide required completeness and validity of estimation; 2) a method of software quality assessment using the classified metrics set, which allows to define model of metric assessment of software quality, to obtain a scalar estimation or to reduce dimension of evaluation vector, and also to analyze goodness of the software metric assessment; 3) a method of the software reliability assessment, which allows to realize systematic choice and verification of software reliability models on the basis of the formalized analysis of assumptions about processes and products of software development. The developed methods and tools allow to improve completeness and validity of estimation of software quality and reliability characteristics and verification of requirements implementation taking into account possible resources limitations, and also to reduce a part of manual operations during software quality estimation