WEMA versus B-WEMA Methods in Forex Forecasting

Hansun, Seng (2017) WEMA versus B-WEMA Methods in Forex Forecasting. ICMLC 2017: Proceedings of the 9th International Conference on Machine Learning and Computing.

Full text not available from this repository.

Abstract

Weighted Exponential Moving Average (WEMA) method is a new hybrid moving average method which combines the weighting factor calculation found in Weighted Moving Average method with Exponential Moving Average method. It had been proven on previous study that the method can give a better accuracy and robustness levels compared to other conventional moving average methods. Another study which combined the Weighted Moving Average method with Brown's Double Exponential Smoothing method had also been done. The proposed method is known as Brown's Weighted Exponential Moving Average (B-WEMA) method and had been proven to excel other conventional moving average methods in terms of the accuracy and robustness levels. In this study, we will try to compare WEMA and B-WEMA forecasting methods in time series analysis, especially in forecasting. We will implement both methods to forecast three major foreign exchange (forex) data transactions and compare the performance of both methods by using Mean Square Error and Mean Absolute Percentage Error criteria. From the experiments taken, it can be concluded that WEMA and B-WEMA have quite the same accuracy and robustness levels due to their slightly same MSE and MAPE values.

Item Type: Article
Subjects: 000 Computer Science, Information and General Works > 000 Computer Science, Knowledge and Systems > 006 Special Computer Methods
300 Social Sciences > 330 Economics > 332 Financial Economics (Shares, Investment)
Divisions: Faculty of Engineering & Informatics > Informatics
Depositing User: Administrator UMN Library
Date Deposited: 21 Oct 2021 01:24
Last Modified: 21 Oct 2021 01:24
URI: https://kc.umn.ac.id/id/eprint/18892

Actions (login required)

View Item View Item