Indian Journal of Science and Technology

Abstract

Objectives: Enhancement of the effort estimation by using Fuzzy Logic with COCOMO-II Effort Multiplier and comparing BRE, RE, VAF. Methods/Statistical Analysis: For the cost estimation, fuzzy logic is used with three membership functions such as Triangular, Trepezoidal and Bell. Findings: The results show that all three membership functions vary by 1-2 % in the effort estimation where as if compared; COCOMO 2 Effort varies by 5-10 % from Fuzzy output. The traditional method consumes a lot of time and a lot of methods are non mathematical due to which the predicted results may be irrelevant. Application/Improvements: The work has been tested on 5 projects. When the parameters for triangular function are executed, values for embedded 1, semidetached 2 are enhanced as compare to organic, semi detached 1, embedded 2. After the evaluation of trapezoidal function, values of semidetached 1 and embedded 1 are more as compare to others. Same as before, the values for embedded 1 and semidetached 2 are more in Bell membership function. Slight improvement has been seen in organic value. Similarly, for COCOMO II the results have been obtained.

Keywords: COCOMO-II, Cost Metrics, Fuzzy Logic, Software Effort Estimation