We investigate the generalization properties of a data-mining approach to single-position day trading which uses an evolutionary algorithm to construct fuzzy predictive models of financial instruments. The models, expressed as fuzzy rule bases, take a number of popular technical indicators on day t as inputs and produce a trading signal for day t + 1 based on a dataset of past observations of which actions would have been most profitable. The approach has been applied to trading several financial instruments (large-cap stocks and indices), in order to study the horizontal, i.e., cross-market, generalization capabilities of the models.
Horizontal generalization properties of fuzzy rule-based trading models / C. da Costa Pereira, A.G.B. Tettamanzi - In: Applications of evolutionary computing : EvoWorkshops 2008 : EvoCOMNET, EvoFIN, EvoHOT, EvoIASP, EvoMUSART, EvoNUM, EvoSTOC, and EvoTransLog : Naples, Italy, march 26-28, 2008 : proceedings / [a cura di] Mario Giacobini ... [et al.]. - Berlin : Springer, 2008. - ISBN 9783540787600. - pp. 93-102 (( Intervento presentato al 2. convegno EvoFIN (European Workshop on Evolutionary Computation in Finance and Economics) tenutosi a Naples, Italy nel 2008 [10.1007/978-3-540-78761-7_10].
Horizontal generalization properties of fuzzy rule-based trading models
C. da Costa PereiraPrimo
;A.G.B. TettamanziUltimo
2008
Abstract
We investigate the generalization properties of a data-mining approach to single-position day trading which uses an evolutionary algorithm to construct fuzzy predictive models of financial instruments. The models, expressed as fuzzy rule bases, take a number of popular technical indicators on day t as inputs and produce a trading signal for day t + 1 based on a dataset of past observations of which actions would have been most profitable. The approach has been applied to trading several financial instruments (large-cap stocks and indices), in order to study the horizontal, i.e., cross-market, generalization capabilities of the models.
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
