Fuzzy-Evolutionary Modeling for Single-Position Day Trading

摘要: This chapter illustrates a data-mining approach to single-position day trading which uses an evolutionary algorithm construct fuzzy predictive model of financial instrument. The is expressed as set IF-THEN rules. takes inputs the open, high, low, and close prices, well values number popular technical indicators on t produces go short, do nothing, long signal for t+1 based dataset past observations actions would have been most profitable. has applied several instruments (large-cap stocks indices): experimental results are presented discussed. A method enhance performance rules by using ensembles models finally illustrated. clearly indicate that, despite its simplicity, may yield significant returns, outperforming buy-and-hold strategy.