An Empirical Model of Fault Prediction in Software Engineering

Abstract: Even though various approaches developed by the many researchers, they may not be optimal while predication of faults. In this approach we are introducing the fault prediction approach with OO metrics along with cyclomatic complexity and nested block depth, in acceptance testing, each function specified in the design document can be independently tested, that is, a set of test cases is developed for each function, not for each workflow module or other module/component. Our experimental results show the efficient fault prediction with our computation parameters.Our approach mainly concentrates on the count of faults prior to testing, expected number of faults, our classification which involves algorithmic and processing, control, logic and sequence, typographical Syntax errors i.e. incorrect spelling of a variable name, regular iteration of statements, incorrect initialization statements per module, this proposed classification approach shows optimal results while analyzing the metrics with training samples after measurement