Prediction of bullish and bearish candlestick signals movement on forex using random forest and multilayer perceptron

DA, ZEBE and Halim, Siana (2024) Prediction of bullish and bearish candlestick signals movement on forex using random forest and multilayer perceptron. [UNSPECIFIED]

	PDF Download (686Kb)
	PDF Download (1410Kb)
Preview	PDF (paper - Siana H) Download (4Mb) | Preview

Abstract

This paper discusses the bullish and bearish candlestick signals using random forest (RF) and multilayer perceptron (MLP). We have two scenarios to apply. First, we used the stochastics measurements as the features of the RF and MLP. Second, we added the candlestick features into the models. In the first scenario, the accuracy rate for the random forest is 61.68%, while the MLP gets an accuracy of 64.15%. Adding the candlestick features increases the accuracy of the prediction both for the RF and the MLP. In the second scenario, the random forest improved up to 82.08%, and MLP gained 72.88% accuracy.

Actions (login required)