本文選題：螞蟻算法 + 元胞自動機��； 參考：《上海金融》2017年01期

【摘要】：蟻群優(yōu)化算法是一種新型的解決組合優(yōu)化問題的仿真型算法,在許多優(yōu)化計算領域中都有廣泛的應用,但卻有容易陷入局部最優(yōu)等方面的缺陷。本人之前研究將元胞自動機原理引入螞蟻算法,構造元胞螞蟻算法,該算法從理論上被證明能部分避免螞蟻算法的缺陷從而提高算法的有效性。本文針對商業(yè)銀行信用風險評估模型的特性,對元胞螞蟻算法的選擇策略、元胞機制、信息素更新機制及尋優(yōu)終止條件等機制進行改造,提出一種具備自我學習和系統(tǒng)信息反饋機制的商業(yè)銀行信用風險評估元胞螞蟻算法模型。最后用DELPHI做了算法仿真實驗,結果驗證了該模型在信用風險評估中具有較好準確性。
[Abstract]:Ant colony optimization (ACO) algorithm is a new simulation algorithm for combinatorial optimization problems. It is widely used in many fields of optimization, but it is easy to fall into local optimization. In this paper, the principle of cellular automata is introduced into ant algorithm to construct a cellular ant algorithm. This algorithm has been proved theoretically to partially avoid the defect of ant algorithm and improve the effectiveness of the algorithm. In this paper, according to the characteristics of credit risk assessment model of commercial banks, the selection strategy, cellular mechanism, pheromone updating mechanism and optimization termination condition of the Cellular Ant algorithm are modified. A cellular ant algorithm model for credit risk assessment of commercial banks with self-learning and systematic information feedback mechanism is proposed. Finally, the algorithm is simulated with Delphi, and the results show that the model is accurate in credit risk assessment.
【作者單位】： 復旦大學應用經(jīng)濟學流動站;上海市高級人民法院;
【分類號】：F832.4
，

本文編號：2063432