Modelling And Forecasting Monthly Tourist Arrivals To United States And Indonesia Using Arima Hybrids Of Machine Learning And Statistical Methods To Assess The Impact Of COVID-19

Misengo, Edward Exavery (2021) Modelling And Forecasting Monthly Tourist Arrivals To United States And Indonesia Using Arima Hybrids Of Machine Learning And Statistical Methods To Assess The Impact Of COVID-19. Masters thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 06211950017011_Edward_Exavery_Misengo_Thesis_Final_(6_8_2021).pdf] Text
06211950017011_Edward_Exavery_Misengo_Thesis_Final_(6_8_2021).pdf - Accepted Version
Restricted to Repository staff only until 1 October 2023.

Download (6MB)
[thumbnail of 06211950017011_Edward_Exavery_Misengo_Thesis_Final_(6_8_2021).pdf] Text
06211950017011_Edward_Exavery_Misengo_Thesis_Final_(6_8_2021).pdf
Restricted to Repository staff only

Download (6MB)
Official URL: https://icosmee-uns.org

Abstract

Tourism is the one of the key economic sectors which contributes significantly to the values of Gross Domestic Product (GDP) for both developed and developing countries. Various sectors in different parts of the world have been affected by Corona Virus Disease (COVID-19) including tourism sector since its outbreak in Wuhan, China on late December 2019. This situation has led to the decrease in number of tourist arrivals to the United States and Indonesia because of some measures and restrictions most of the countries have been taken to combat and reduce the spread of the disease. This study devotes more on modelling and forecasting tourist arrivals to the United States and Indonesia using both ARIMA hybrids of statistical methods (TSMR & HW) and ARIMA hybrids of machine learning methods (MLP & LSTM). Based on MAPE values, hybrid models involving ARIMA and MLP models are observed to perform better in forecasting monthly tourist arrivals to both United States and Indonesia than hybrid models involving ARIMA with HW, TSMR and LSTM especially when ARIMA used as auxiliary forecasting model.
Hybrid models "MLP" (6,1)-〖"ARIMA" (0,1,1)(0,1,1)〗_12 (RMSE=211,837.64 & MAPE=2.79%) and "MLP" (12,1)-〖"ARIMA" (0,1,1)(0,1,0)〗_12 (RMSE=88,636.87 & MAPE=4.92%) as the best hybrid models have been used in this study to forecast monthly tourist arrivals to the United States and Indonesia respectively in the period of 2018-2019 and in the year 2020. The forecasts obtained from the selected best hybrid models are compared to actual monthly tourist arrivals recorded in year 2020 to assess the impact of COVID-19. The finding shows that, the decrease in the number of foreign visitors because of COVID-19 the United States and Indonesia lost about 118,891,084,736 (US $) and 14,217,637,566.96 (US $) respectively as expected revenues due to the estimated expenditure of foreign visitors in 2020.
=====================================================================================================

Item Type: Thesis (Masters)
Uncontrolled Keywords: Autoregressive Integrated Moving Average (ARIMA), Holt-Winters Exponential Smoothing Method (HW), Long Short Term Memory (LSTM), Multilayer Perceptron (MLP),Time Series Multiple Regression (TSMR).
Subjects: H Social Sciences > HA Statistics > HA30.3 Time-series analysis
Q Science > Q Science (General) > Q325.5 Machine learning.
R Medicine > RA Public aspects of medicine > RA644.C67 COVID-19 (Disease)
Divisions: Faculty of Mathematics and Science > Statistics > 49101-(S2) Master Thesis
Depositing User: Edward Exavery Misengo
Date Deposited: 06 Aug 2021 08:39
Last Modified: 06 Aug 2021 08:39
URI: http://repository.its.ac.id/id/eprint/85061

Actions (login required)

View Item View Item