Portfolio Selection Using Neural Network
Portfolio Selection Using Neural Network
seyed mohsen pilehvar1
1) Master financial engineering emai:
IOI Code
:
XYBB-ZHEZH
Publication :
3rd International Conference On Research Science And Technology(3rstconf.com)
Abstract :
In this paper we trying to solve generalization of standard Markowitz mean variance model which include cardinality and bounding constraints using Hopfield neural network.
We apply new model for present new efficient frontier that in addition to show minimum amount of risk for a given level of return, this calculate minimum Value at Risk (VaR) too.
All of these evaluate using Hopfield neural network and other heuristic algorithms (GA & PSO) by foreign and Iranian data. The foreign data correspond to weekly prices from March 1992 to September 1997 and they come from the indices HangSeng and we consider weekly prices between 1385 to 1390 from 40 stock of 50 that were be better(50 better corporation) in year 1390 in Tehran stock exchange(TSE).
Keywords : portfolio selection hopfield neural network tehran stock exchange