Time series modeling of international visitor arrivals in Tanzania
| dc.contributor.author | Mwiru, Paskasi Dominick | |
| dc.date.accessioned | 2020-06-04T12:20:01Z | |
| dc.date.available | 2020-06-04T12:20:01Z | |
| dc.date.issued | 1999 | |
| dc.description | Available in print form, East Africana Collection, Dr. Wilbert Chagula Library, Class Mark ( THS EAF HA30.3.M854) | en_US |
| dc.description.abstract | This study looked at the ten-year trend behavior and the market share of international visitor arrivals in Tanzania. Issues associated with forecasting univariate time series of international visitor arrivals in Tanzania were explored using Box and Jenkins methodology. Annual data from 1995 to 2004 of international visitor arrivals by nationality and monthly data from January 1995 to December 2006 of international visitor arrivals are used to analyze the market share and model fitting respectively. Using Microsoft Excel the market behavior trends and market share of the market sources were determined. Line graphs and Pie charts were in determining and presenting some of the trends and market shares. The trend and market shares were included in the discussion to determine the main source markets of international visitor arrivals in Tanzania. For the case of time domain analysis of monthly international visitor arrivals in Tanzania several transformation were made to make the data stationary. These included logarithmic transformation and differencing monthly international visitor arrivals other Box and Jenkins stages were also adopted. In identification stage, monthly international visitor arrivals’ autocorrelations were plotted and examined as well. Four candidate ARIMA models were selected basing on Akaike Information Criterion (AIC), Ljung-Box Q and Durbin Warson statistics Root Mean Square error (RMSE) and Mean percentage error (PME). ARIMA (4,1,3) model outperformed other ARIMA (p,d,q) models. Seasonal models were also considered and ARIMA (1,0,1)×(2,1,2)12 outperformed other SARIMA models in terms of AIC, RMSE and MPE. But ARIMA (4,1,3) outperformed this model in terms of these criteria. Both ARIMA (4,1,3) and SARIMA (1,0,1)×(2,1,2)12 were used in forecasting international visitor arrivals and ARIMA (4,1,3) produced better forecasts as compared to the SARIMA model. | en_US |
| dc.identifier.citation | Mwiru, P. D (1999) Time series modeling of international visitor arrivals in Tanzania, Master dissertation, University of Dar es Salaam | en_US |
| dc.identifier.uri | http://41.86.178.5:8080/xmlui/handle/123456789/12044 | |
| dc.language.iso | en | en_US |
| dc.publisher | University of Dar es Salaam | en_US |
| dc.subject | Time-series analysis | en_US |
| dc.subject | Visitors | en_US |
| dc.subject | Foreign | en_US |
| dc.subject | Tanzania | en_US |
| dc.title | Time series modeling of international visitor arrivals in Tanzania | en_US |
| dc.type | Thesis | en_US |