Robust Rank Aggregation method for Case-Base effort estimation
Ali Bou Nassif, Mohammad Azzeh, Shadi Banitaan · 2017
It is well recognized that effort estimation is an essential part of successful software management. Among many estimation models, the Case-Base Effort Estimation (CBEE) has been intensively used among researchers and practitioners as a promising model for better and accurate effort prediction. The common challenges with this model are: (1) finding the nearest cases to the new case, (2) selecting best features set, and (3) determining the optimal k number of nearest cases that should be involved in the final prediction stage. Irrespective of existing many models to mitigate these challenges, none have achieved the full trust from research community. In this paper, we used Robust Rank Aggregation method as a potential solution for the abovementioned challenges. We believe that the CBEE tasks can been seen as ranking problem. The proposed model has been evaluated over 8 datasets coming from different sources. The results achieved are promising. In conclusion, we can recommend the proposed model as a successful alternative for original CBEE.