Research on complex financial decision making driven by geometric aggregation and intelligent optimization of high-dimensional expert information.

In complex fields such as finance, multi-attribute group decision-making (MAGDM) often faces challenges such as high-dimensional expert opinions, fuzzy uncertainty, and information heterogeneity. To address these issues, this paper proposes a high-dimensional intelligent aggregation framework that integrates Picture Fuzzy Z-numbers with the Plant Growth Simulation Algorithm (PGSA). By employing a biomimetic phototaxis-based search mechanism, the framework dynamically identifies the spatially optimal aggregation points within the expert preference point set. Using the Ashraf financial decision-making dataset as an example, this paper systematically compares the proposed method with seven mainstream aggregation techniques. Experimental results demonstrate that the proposed method exhibits significant advantages in key metrics: Hamming distance of 0.0928, weight cosine Similarity of 0.9793, information energy of 0.1138, and Pearson correlation coefficient of 0.9277, outperforming most existing methods and demonstrating higher aggregation accuracy. This study provides new insights into fuzzy information integration in complex financial decision-making and offers an effective tool for high-dimensional group preference modeling and aggregation optimization.