Post-evaluation of water allocation based on non-negative matrix factorization

Based on establishing an evaluation index system, this paper developed a new post-evaluation model based on non-negative matrix factorization (NMF) for diagnostic evaluation. The model uses NMF to decompose the data matrix in to the base vector and the weight vector to the largest approximation degree. Then the base vector is taken as an assessment basis of the evaluation index, and the weight vector is used for grading the evaluation scheme. This model is applied to evaluate the water allocation scheme for Luanhe River from perspectives of scheme quality, implementation process, implementation effect and sustainability. The overall evaluation grade of the scheme is excellent, but there is still room for improvement regarding water sustainability, impact on the water environment, guarantee rate of water quality targets and so on. Results show that the post-evaluation model based on NMF can produce satisfactory outcomes of the post-evaluation considering the system as a whole.

Project Supported:

Clc Number: