Water Eutrophication Evaluation Based on the Improved Projection Pursuit Regression Model Under the Hesitant Fuzzy Environment

Abstract

The water eutrophication restricts the development of economy and society in China, which attracts increasing attention. It also affects the health and ecological environment. The evaluation of water eutrophication is very complicated due to the dynamic variability of the water quality data. This paper adopts the hesitant fuzzy set (HFS) to depict the massive data of samples and uncertain preference information of experts, which reduces the complexity of calculation and avoids the loss of information. After that, we construct the projection index function based on the main factors of water eutrophication. The particle swarm optimization (PSO) algorithm is applied to determine the global optimal projection direction by optimizing the projection index function. Therefore, we construct an improved projection pursuit regression (PPR) model. Finally, the water eutrophication evaluation of several lakes in China is used to demonstrate the improved PPR model. Also, the comparative analysis and contribution rate analysis are conducted to validate its rationality and advantages.