Atlantis Press Paper Details: Multi-Layer Allocated Learning Based Neural Network for Resource Allocation Optimization

The investor’s asset allocation choice deeply depends on the trade-off between risk and return. The well-known mean variance method requires predetermined risk and expected return to calculate optimal investment weights of portfolio. The artificial neural network (ANN) with nonlinear capability is proven to solve large-scale complex problem effectively. However, the traditional ANN model cannot guarantee to produce reasonable investment allocation, because the summation of investment weight may not preserve 100% in output layer. This article introduces a multi-layer allocated learning based neural network model and takes financial portfolio as an example to optimize assets allocation weights. This model dynamically adjusts the investment weight as a basis of 100% of summing all of asset weights in the portfolio. The experimental results demonstrate the feasibility of optimal investment weights and superiority of ROI of buy-and-hold trading strategy compared with benchmark TSE (Taiwan Stock Exchange).


title:		Multi-Layer Allocated Learning Based Neural Network for Resource Allocation Optimization
publication:		JCIS-2006 Proceedings
part of series:		Advances in Intelligent Systems Research
ISBN:		978-90-78677-01-7
ISSN:		1951-6851
DOI:		doi:10.2991/jcis.2006.201 (how to use a DOI)
author(s):		Po-Chang Ko, Ping-Chen Lin, Jan-An You, Yu-Jen Tien
corresponding author:		Po-Chang Ko
publication date:		October 2006
keywords:		Resource allocation, neural network, allocated learning based NN, portfolio, investment weight
abstract:		The investor’s asset allocation choice deeply depends on the trade-off between risk and return. The well-known mean variance method requires predetermined risk and expected return to calculate optimal investment weights of portfolio. The artificial neural network (ANN) with nonlinear capability is proven to solve large-scale complex problem effectively. However, the traditional ANN model cannot guarantee to produce reasonable investment allocation, because the summation of investment weight may not preserve 100% in output layer. This article introduces a multi-layer allocated learning based neural network model and takes financial portfolio as an example to optimize assets allocation weights. This model dynamically adjusts the investment weight as a basis of 100% of summing all of asset weights in the portfolio. The experimental results demonstrate the feasibility of optimal investment weights and superiority of ROI of buy-and-hold trading strategy compared with benchmark TSE (Taiwan Stock Exchange).
copyright:		© Atlantis Press. This article is distributed under the terms of the Creative Commons Attribution License, which permits non-commercial use, distribution and reproduction in any medium, provided the original work is properly cited.
full text:		JCIS06-CSI-132.pdf (1213 K)