A Review: Effort Estimation Model for Scrum Projects using Supervised Learning

Abstract

Effort estimation practice in Agile is a critical component of the methodology to help cross-functional teams to plan and prioritize their work. Agile approaches have emerged in recent years as a more adaptable means of creating software projects because they consistently produce a workable end product that is developed progressively, preventing projects from failing entirely. Agile software development enables teams to collaborate directly with clients and swiftly adjust to changing requirements. This produces a result that is distinct, gradual, and targeted. It has been noted that the present Scrum estimate approach heavily relies on historical data from previous projects and expert opinion, while existing agile estimation methods like analogy and planning poker become unpredictable in the absence of historical data and experts. User Stories are used to estimate effort in the Agile approach, which has been adopted by 60–70% of the software businesses. This study's goal is to review a variety of strategies and techniques that will be used to gauge and forecast effort. Additionally, the supervised machine learning method most suited for predictive analysis is reviewed in this paper.