Study on the improved grey decision-making model and its application

Existing clustering algorithms need to specify the number of clusters and to select initial points using human input, which lead to inferior clustering and optimisation outputs. Here, an improved grey decision-making model based on the thought of affinity propagation algorithm and grey correlation analysis is proposed to solve these problems. According to the panel data class and the inter-class candidate points between the message dissemination for clustering, we fully mine all information contained in a multi-indicator panel dataset. Finally, a case study is used to test the improved model's validity and rationality.