When coping with complex global optimization problems, often it is not possible to obtain either analytical or exact solutions. Therefore, one is forced to resort to approximate numerical optimizers. With this aim, several metaheuristics have been proposed in the literature and the primary approaches can be traced back to biology and physics. On one hand, there exist bio-inspired metaheuristics that imitate the Darwinian evolution of species (like, for instance, Genetic Algorithms) or the behaviour of group of social organisms (like, for instance, Ant Colony Optimization). On the other hand, there exist physics-inspired metaheuristics that mimic physical laws (like, for instance, gravitation and electromagnetism). In this work, we take into account the Fireworks Algorithm and the Particle Swarm Optimization in order to compare their exploration and exploitation capabilities. In particular, the investigation is performed considering as complex global optimization problem the estimation of the parameters of the technical analysis indicator Bollinger Bands in order to build effective financial trading systems, similarly to what proposed in (Corazza et al., 2019).

When coping with complex global optimization problems, often it is not possible to obtain either analytical or exact solutions. Therefore, one is forced to resort to approximate numerical optimizers. With this aim, several metaheuristics have been proposed in the literature and the primary approaches can be traced back to biology and physics. On one hand, there exist bio-inspired metaheuristics that imitate the Darwinian evolution of species (like, for instance, Genetic Algorithms) or the behaviour of group of social organisms (like, for instance, Ant Colony Optimization). On the other hand, there exist physics-inspired metaheuristics that mimic physical laws (like, for instance, gravitation and electromagnetism). In this work, we take into account the Fireworks Algorithm and the Particle Swarm Optimization in order to compare their exploration and exploitation capabilities. In particular, the investigation is performed considering as complex global optimization problem the estimation of the parameters of the technical analysis indicator Bollinger Bands in order to build effective financial trading systems, similarly to what proposed in[3].

Exploration and Exploitation in Optimizing a Basic Financial Trading System: A Comparison Between FA and PSO Algorithms

Claudio Pizzi
;
Irene Bitto;Marco Corazza
2021-01-01

Abstract

When coping with complex global optimization problems, often it is not possible to obtain either analytical or exact solutions. Therefore, one is forced to resort to approximate numerical optimizers. With this aim, several metaheuristics have been proposed in the literature and the primary approaches can be traced back to biology and physics. On one hand, there exist bio-inspired metaheuristics that imitate the Darwinian evolution of species (like, for instance, Genetic Algorithms) or the behaviour of group of social organisms (like, for instance, Ant Colony Optimization). On the other hand, there exist physics-inspired metaheuristics that mimic physical laws (like, for instance, gravitation and electromagnetism). In this work, we take into account the Fireworks Algorithm and the Particle Swarm Optimization in order to compare their exploration and exploitation capabilities. In particular, the investigation is performed considering as complex global optimization problem the estimation of the parameters of the technical analysis indicator Bollinger Bands in order to build effective financial trading systems, similarly to what proposed in[3].
2021
Progresses in Artificial Intelligence and Neural Systems
File in questo prodotto:
File Dimensione Formato  
Pizzi Bitto Corazza Progress_in_Artificial_Intelligence_and_Neural_Systems-2021.pdf

non disponibili

Tipologia: Versione dell'editore
Licenza: Accesso gratuito (solo visione)
Dimensione 324.54 kB
Formato Adobe PDF
324.54 kB Adobe PDF   Visualizza/Apri

I documenti in ARCA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10278/3727810
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? ND
social impact