In recent decades the production of large software projects are very large and is costly and time consuming during the phases of software development there are some bugs. Some of the errors generated by the software to detect errors in the initial is phases these errors and may not be seen until the final phases. To clear this error may be the next generation of software. Time and expense of producing the software is error. Error in this phase will increase the cost and time. Over time, larger projects And the error in estimating software cost is higher and higher. and these days detecting the possible defect is one of consideration to rely on software quality. So there is a need to create a prediction model and we can use data mining methods to predict defects. This paper examined ways of imposing clustering on various projects and putting them in groups with the similar characteristics. By using this pattern we can choose a defect predication model that is able to predict the defect of whole group.
展开▼
