Adoption of machine learning techniques in software effort estimation: an overview

Abstract

Nowadays the significant trend of the effort estimation is in demand. It needs more data to be collected and the stakeholders require an effective and efficient software for processing, which makes the hardware and software cost development becomes steeply increasing. This scenario is true especially in the area of large industry, as the size of a software project is becoming more complex and bigger, the complexity of estimation is continuously increased. Effort estimation is part of the software engineering economic study on how to manage limited resources in a way a project could meet its target goal in a specified schedule, budget and scope. It is necessary to develop or adopt a useful software development process in executing a software development project by acting as a key constraint to the project. The accuracy of estimation is the main critical evaluation for every study. Recently, the machine learning techniques are becoming widely used in many effort estimation problems but there are limitations in some of the models and the variation research is still not enough. This paper presents an overview of the effort estimation using machine learning techniques and will be useful for researchers to provide future direction in the field of machine learning adoption in software effort estimation.