學術產出-NSC Projects

Article View/Open

Publication Export

Google ScholarTM

政大圖書館

Citation Infomation

  • No doi shows Citation Infomation
題名 運用新的粒子群演算法求解馬可維茲資產組合選擇模型
其他題名 A New Particle Swarm Optimization for Markowitz Portfolio Selection Model
作者 陳春龍
貢獻者 國立政治大學資訊管理學系
行政院國家科學委員會
關鍵詞 粒子群演算法;資產組合;馬可維茲模型;效率前緣
Particle swarm optimization; Portfolio selection; Markowitz mean-variance model; Efficient frontier
日期 2010
上傳時間 30-Aug-2012 15:48:44 (UTC+8)
摘要 粒子群演算法(Particle Swarm Optimization,PSO)是一種以群體為基礎的最佳 化搜尋方法,由Kenney 和Eberhart 於1995 年提出,並已經成功的應用在解決一些困 難的最佳化問題上。在本研究中,我們將嘗試利用一些方法來改善基本粒子群演算法過 早收斂而可能陷入區域最佳解的缺點,進而改善粒子群演算法的績效。我們已經使用常 見的Benchmark 函數來評估新的粒子群演算法之績效。我們將持續改善新的粒子群演算 法並將之應用於求解著名的馬可維茲資產組合選擇模型;我們將採用與Chang et al. 相同的研究實驗數據資料,求解不同條件下資產組合所構成的效率前緣(Efficient Frontier)。最後,我們將比較新的粒子群演算法與基因演算法(Genetic Algorithms)、 模擬退火法(Simulated Annealing)、與禁忌搜尋法(Tabu Search)等方法求解馬可維 茲資產組合選擇模型的績效。
Particle swarm optimization (PSO), introduced by Kennedy and Eberhart in 1995, is a social population-based search algorithm that has been successfully applied to solve various hard optimization problems. In this research, we have tried to develop some ideas to overcome the major drawback of the basic PSO, which is swarm stagnation, in order to improve its performance. We have applied the new PSO to some commonly used benchmark functions and have produced promising results. We will continue to improve the performance of the new PSO and will apply it to solve the well-known Markowitz mean-variance portfolio selection model. We will utilize the model and the benchmark data used in Chang et al. to generate the efficient frontiers under different conditions. We will evaluate the performance of the new PSO by comparing the efficient frontiers produced by the new PSO, genetic algorithms, simulated annealing, and tabu search.
關聯 應用研究
學術補助
研究期間:9908~ 10007
研究經費:536仟元
資料類型 report
dc.contributor 國立政治大學資訊管理學系en_US
dc.contributor 行政院國家科學委員會en_US
dc.creator (作者) 陳春龍zh_TW
dc.date (日期) 2010en_US
dc.date.accessioned 30-Aug-2012 15:48:44 (UTC+8)-
dc.date.available 30-Aug-2012 15:48:44 (UTC+8)-
dc.date.issued (上傳時間) 30-Aug-2012 15:48:44 (UTC+8)-
dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/53419-
dc.description.abstract (摘要) 粒子群演算法(Particle Swarm Optimization,PSO)是一種以群體為基礎的最佳 化搜尋方法,由Kenney 和Eberhart 於1995 年提出,並已經成功的應用在解決一些困 難的最佳化問題上。在本研究中,我們將嘗試利用一些方法來改善基本粒子群演算法過 早收斂而可能陷入區域最佳解的缺點,進而改善粒子群演算法的績效。我們已經使用常 見的Benchmark 函數來評估新的粒子群演算法之績效。我們將持續改善新的粒子群演算 法並將之應用於求解著名的馬可維茲資產組合選擇模型;我們將採用與Chang et al. 相同的研究實驗數據資料,求解不同條件下資產組合所構成的效率前緣(Efficient Frontier)。最後,我們將比較新的粒子群演算法與基因演算法(Genetic Algorithms)、 模擬退火法(Simulated Annealing)、與禁忌搜尋法(Tabu Search)等方法求解馬可維 茲資產組合選擇模型的績效。en_US
dc.description.abstract (摘要) Particle swarm optimization (PSO), introduced by Kennedy and Eberhart in 1995, is a social population-based search algorithm that has been successfully applied to solve various hard optimization problems. In this research, we have tried to develop some ideas to overcome the major drawback of the basic PSO, which is swarm stagnation, in order to improve its performance. We have applied the new PSO to some commonly used benchmark functions and have produced promising results. We will continue to improve the performance of the new PSO and will apply it to solve the well-known Markowitz mean-variance portfolio selection model. We will utilize the model and the benchmark data used in Chang et al. to generate the efficient frontiers under different conditions. We will evaluate the performance of the new PSO by comparing the efficient frontiers produced by the new PSO, genetic algorithms, simulated annealing, and tabu search.en_US
dc.language.iso en_US-
dc.relation (關聯) 應用研究en_US
dc.relation (關聯) 學術補助en_US
dc.relation (關聯) 研究期間:9908~ 10007en_US
dc.relation (關聯) 研究經費:536仟元en_US
dc.subject (關鍵詞) 粒子群演算法;資產組合;馬可維茲模型;效率前緣en_US
dc.subject (關鍵詞) Particle swarm optimization; Portfolio selection; Markowitz mean-variance model; Efficient frontieren_US
dc.title (題名) 運用新的粒子群演算法求解馬可維茲資產組合選擇模型zh_TW
dc.title.alternative (其他題名) A New Particle Swarm Optimization for Markowitz Portfolio Selection Modelen_US
dc.type (資料類型) reporten