The cost-effectiveness evaluation of marketing and maintenance projects by the RankNet-assisted algorithm

The cost-effectiveness evaluation for power enterprises is concentrated in the field of power grid construction projects with large investment scales, while such analysis aiming at marketing and maintenance projects (MMPs) remains to be studied. Meanwhile, issues such as the low universality of evaluation indices and the over-complexity of the evaluation model culminate in the gap between the theory and practical application. This paper proposes a novel RankNet-based algorithm to prioritize MMPs, which provides a scientific reflection of both professional and economic impacts, thereby addressing the shortcomings in current cost-effectiveness evaluation. Firstly, an unsupervised learning technique is adopted to select representative projects of different types. Secondly, a comprehensive evaluation system with reasonable weights for each index is established by the Analytic Hierarchy Process and Entropy Weight Method (AHP-EWM). On this basis, a fuzzy comprehensive evaluation is presented to generate the scores thus the cost-effectiveness of each project is quantified. Finally, the advantages of the proposed method are verified compared with traditional methods by a practical case.