A Comparative Study of Algorithms of Software Effort Estimation for the Robotic and Communication Systems Based on Improved Accuracy

Abstract

The engineering systems of robotics, communication networks, and electronics status, require a software effort estimation to decrease the error of effort and cost estimation since huge sizes of datasets are used in these systems. It supports the different tasks in scheduling, planning, and so on yet it is difficult to estimate the necessary duration to fix a required task. However, the computational complexity level is increased with improving of abovementioned systems. In this paper, several software effort estimation techniques are considered for mechatronics and communications systems. These techniques are artificial neural networks, Fuzzy logic rule, genetic algorithm, and others. The analyses and investigations revealed that the hybrid technique is the best one, which is described as the statistical representations cascaded to artificial neural networks. the hybrid technique has a higher accuracy with desirable complexity.