基于敏捷开发的工作量估算模型的研究

Research of effort estimation model based on agile development

为了能够对敏捷开发项目进行有效的工作量估算,提出了一种基于人工神经网络的工作量估算模型。根据ISBSG(international software benchmarking standard group)数据集的特点,对数据集进行预处理。利用主成分分析方法对工作量的影响因子进行压缩,得到了针对敏捷开发的综合影响因子集合。基于BP神经网络,建立了工作量估算模型。给出了该模型的具体实现步骤和实验验证,并将实验结果与其他估算方法进行比较。实验结果表明,该模型能很好地实现敏捷开发的工作量估算,且明显优于其他估算方法。

To estimate the effort of agile development projects effectively, the effort estimation model is proposed based on artifi cial neural network. Firstly, according to the characteristics of ISBSG data set, the data set is processed. Secondly, the influen cing factors are compressed, and the influencing factors set are obtained. Then, the effort estimation model is established based on BP neural network. The implementation steps and experimental verification are given for the model, and this model is com pared with other estimation methods. Experiments show that this model has a good performance on the effort estimation for agile development, and is better than other estimation methods.