COCOMO II IMPROVEMENT USING K-MEANS TECHNIQUE TO INCREASE THE ACCURACY OF EFFORT ESTIMATION

• Main Author: Rudiyanto, Chintia
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/36802
• Institution: Institut Teknologi Bandung

Effort estimation is one of the important processes in a software project. There are several techniques and models that can be used to estimate effort, one of which is COCOMO II. COCOMO II is a development of COCOMO '81 and is still developing today. COCOMO II is quite widely used because of its open source. Everyone can see and make use of the COCOMO II effort estimation model. However, the accuracy of the COCOMO II model is considered to be lacking and needs to be improved. In this study, a model that can increase the accuracy value of COCOMO II is proposed by utilizing the K-Means clustering technique. In the proposed model, the K-Means clustering technique is used to determine the training data that will be used in the COCOMO II calibration process. The COCOMO II calibration aims to determine the new A and B constant values. So that in the proposed model, each project estimated will have different values of A and B. Based on the results of the study, it can be concluded that the accuracy of the proposed model is generally increased compared to the original COCOMO II. The value of accuracy depends on the preprocessing technique performed and the number of clusters. For COCOMO NASA 2 datasets and Turkish Software Industry, the best accuracy is achieved when the preprocessing technique performed is to give 100 weight to all cost driver attributes and the number of clusters is 5. This proposed model can reduce the MRE COCOMO II value from 1.32 to 0.85 and increase the value of PRED (0.3) from 32% to 54%. PRED value (0.3) = 54% means that the proposed model can estimate approximately 30% of the actual for as many as 54% of projects estimated.