Until now, numerous effort estimation models for software projects have been developed, most of them producing accurate results but not providing the flexibility to decision makers during the software development process. The main objective of this study is to objectively and accurately estimate the effort when using the Scrum methodology. A dynamic effort estimation model is developed by using regression-based machine learning algorithms. Story point as a unit of measure is used for estimating the effort involved in an issue. Projects are divided into phases and the phases are respectively divided into iterations and issues. Effort estimation is performed for each issue, then the total effort is calculated with aggregate functions respectively for iteration, phase and project. This architecture of our model provides flexibility to decision makers in any case of deviation from the project plan. An empirical evaluation demonstrates that the error rate of our story point-based estimation model is better than others.

中文翻译：

---

使用机器学习技术的基于故事点的努力估计模型

到目前为止，已经开发了许多软件项目的工作量估计模型，其中大多数产生了准确的结果，但在软件开发过程中没有为决策者提供灵活性。本研究的主要目的是客观和准确地估计使用 Scrum 方法时的工作量。通过使用基于回归的机器学习算法开发了动态努力估计模型。故事点作为度量单位用于估计问题所涉及的工作量。项目分为阶段，阶段分别分为迭代和问题。对每个问题进行工作量估计，然后用聚合函数分别计算迭代、阶段和项目的总工作量。我们模型的这种架构在任何偏离项目计划的情况下都为决策者提供了灵活性。实证评估表明，我们基于故事点的估计模型的错误率优于其他模型。