This work is aimed to provide a comprehensive simulation study of dimension reduction techniques to support the decision process underlying the management of financial portfolios. We will test whether it is possible to built allocation schemes in different ways than picking up assets according to classical mean–variance criteria. To such aim, we will examine various dimension reduction techniques and we will merge them into a framework that selects most promising assets according to their proximity to the centroids of each procedure. Extensive simulations will be provided, where we will analyze the performance of the examined techniques, and we will compare them to those of traditional allocation schemes.
Portfolio Optimization with neural dimension reduction techniques: a comprehensive simulation study
RESTA, MARINA
2012-01-01
Abstract
This work is aimed to provide a comprehensive simulation study of dimension reduction techniques to support the decision process underlying the management of financial portfolios. We will test whether it is possible to built allocation schemes in different ways than picking up assets according to classical mean–variance criteria. To such aim, we will examine various dimension reduction techniques and we will merge them into a framework that selects most promising assets according to their proximity to the centroids of each procedure. Extensive simulations will be provided, where we will analyze the performance of the examined techniques, and we will compare them to those of traditional allocation schemes.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
