Adaptive Neuro-Fuzzy Inference System (ANFIS) Based Software Evaluation

Software metric is a measure of some property of a piece of software or its specifications. The goal is to obtain reproducible and quantifiable measurements, which may have several valuable applications in schedule and budget planning, effort and cost evaluation, quality assurance testing, software debugging, software performance optimization, and optimal personnel task assignments. Software effort evaluation is one of the most essential and crucial part of software project planning for which efficient effort metrics is required. Software effort evaluation is followed by software cost evaluation which is helpful for both customers and developers. Thus, efficiency of effort component of software is very essential. The algorithmic models are weak in estimating early effort evaluation with regards to uncertainty and imprecision in software projects. To overcome this problem, there are various machine learning methods. One of the methods is soft computing in which there are various methodologies viz., Artificial Neural Network, Fuzzy Logic, Evolutionary computation based Genetic Algorithm and Metaheuristic based Particle Swarm Optimization. These methods are good at solving real-world ambiguities. This paper highlights the design of an efficient software effort evaluation model using Adaptive Neuro-Fuzzy Inference System (ANFIS) for uncertain datasets and it shows that this technique significantly outperforms with sufficient results.