Abstract: Portfolio selection with probability criterion under investor's preference is the deepening of portfolio selection. It is an optimization mathematic model with the objective function based on probability criterion and investor's preference under the condition of non-negative constrains. A gravitational search algorithm is designed to solve this model. The mass and position correspond to the objective function and investment proportional coefficient, respectively. The implementation of this method is given based on velocity and position update equations Experimental results and comparison experiments show the effectiveness and feasibility of proposed algorithm.

Key words: probability criterion, investor's preference, gravitational search algorithm, optimization

摘要： 投资者偏好条件下概率准则投资组合问题是经典投资组合问题的深化,是在非负投资比例系数约束条件下,以概率准则和偏好因子为目标函数的优化模型。设计了一种基于万有引力搜索算法的求解方法,个体的质量对应于目标函数值,位置对应于投资比例系数,结合速度和位置更新策略,建立了算法的求解步骤。通过数值实验和与现有解法的比较,验证了算法的可行性和有效性。

关键词: 概率准则, 投资者偏好, 万有引力搜索算法, 优化