UMN Knowledge Center

A comparative study on WEMA and H-WEMA forecasting methods in time series analysis (case study: JKSE composite index data)

Hansun, Seng (2017) A comparative study on WEMA and H-WEMA forecasting methods in time series analysis (case study: JKSE composite index data). 2016 6th International Annual Engineering Seminar (InAES).

Full text not available from this repository.
Official URL: https://ieeexplore.ieee.org/document/7821897

Abstract

Moving average as one of popular technical indicator used to predict data in time series analysis has grown significantly. There are many researchers who have develop moving average methods resulting in its' many derivatives methods. Two of them are Weighted Exponential Moving Average (WEMA) and Holt's Weighted Exponential Moving Average (H-WEMA) methods. This research aims to conduct a comparative study on WEMA and H-WEMA methods which are said to excel the other conventional moving average methods. Therefore, we will implement both methods to predict Jakarta Stock Exchange (JKSE) composite index data and then calculate the accuracy and robustness level using Mean Square Error (MSE) and Mean Absolute Percentage Error (MAPE) criteria. The results from the experiments taken show that H-WEMA has a better accuracy and robustness levels compared to other moving average methods.

Item Type: Article
Subjects: 500 Science and Mathematic > 510 Mathematics > 510 Mathematics
500 Science and Mathematic > 510 Mathematics > 519 Probabilities and Applied Mathematics
Divisions: Fakultas Teknik Informatika > Program Studi Informatika
Depositing User: mr admin umn
Date Deposited: 19 Oct 2021 08:37
Last Modified: 19 Oct 2021 08:37
URI: http://kc.umn.ac.id/id/eprint/18889

Actions (login required)

View Item View Item