Optimising Directional Changes trading strategies with different algorithms

Directional Changes (DC), a novel approach for sampling market data, allows the extraction of trends in financial time series by converting series from a time based format to an event-driven format. This paradigm has been shown to give some predictability in financial prediction, and has been used to generate profitable trading strategies on the FOREX market. In the past, a genetic algorithm was used to optimise the parameters of DC-based trading strategy. The goal of this work is to explore whereas different machine learning algorithms can be used to improve the results on the aforementioned optimisation task. For this purpose, we explore two algorithms, namely Particle Swarm Optimization and Shuffled Frog Leaping Algorithm. After comparing the performance of these two algorithms on 36 different datasets from 4 different currency pairs, we find that they statistically improve the profitability of the DC-based trading strategies.