SOFTWARE ENGINEERING AND NEURAL NETWORKS

Abstract

The central theme of this work is the integration of software engineering and artificial neural networks, addressing how the combination of these two areas can result in more efficient, intelligent and adaptive computing solutions. The main objective of the research was to investigate how neural networks, inspired by the functioning of the human brain, can be applied in the context of software development to improve processes such as cost estimation, fault identification and performance optimization. To this end, a bibliographic research methodology was used, based on authors such as Haykin (2001), Pressman (2010), Pfleeger (1998), Silva, Spatti and Flauzino (2010), among others, as well as the analysis of case studies and practical applications already developed in the area. Software engineering was analyzed from its foundations, including its essential phases such as requirements gathering, modeling, coding, testing and maintenance. In parallel, the characteristics of artificial neural networks were explored, such as their ability to learn, adapt and generalize, with emphasis on concrete applications in estimating software development effort. The results of the analysis indicated that incorporating neural networks into the software life cycle can make processes more precise, agile and responsive to change. It was also found that the use of ANNs promotes the automation of tasks that require subjective judgment or specialized knowledge, contributing to the reduction of human failures and improving the final quality of the software product. It is therefore concluded that the convergence between software engineering and artificial neural networks represents a promising trend, capable of transforming the way computer systems are designed, developed and maintained, strengthening technological innovation in the field of computational intelligence.