The combination of interval forecasts in tourism

Combination is an effective way to improve tourism forecasting accuracy. However, empirical evidence is limited to point forecasts. Given that interval forecasts can provide more comprehensive information, it is important to consider both point and interval forecasts for decision-making. Using Hong Kong tourism demand as an empirical case, this study is the first to examine if and how the combination can improve interval forecasting accuracy for tourism demand. Winkler scores are employed to measure interval forecasting performance. Empirical results show that combination improves the accuracy of tourism interval forecasting for different forecasting horizons. The findings provide government and industry practitioners with guidelines for producing accurate interval forecasts that benefit their policy-making for a wide array of applications in practice.This article also launches the Annals of Tourism Research Curated Collection on Tourism Demand Forecast, a special selection of research in this field.     

Tourism forecast combination using the CUSUM technique

Forecast combination in tourism has emerged as an important research area due to its relevance to tourism decision making. This paper further investigates the impact of forecast combination on forecast accuracy by applying a quadratic programming approach to determine the combination weights for individual forecasts. In particular, we introduce three novel ideas which have not been found in previous tourism forecasting studies. First, we introduce a quality control technique, CUSUM, to determine the time for updating the weights. Next, we develop a hybrid method (using quadratic programming) to combine the forecasts to reduce forecasting errors. Thirdly, we investigate whether different performance measures yield different results. Thus, instead of comparing different weighting methods using only one performance measure, we introduce several indicators for performance comparisons. The empirical results suggest that the controlled weighting method both saves time in updating the combination weights and improves the overall performance of the combined forecasts. The method is also easy-to-implement and should be used to improve forecasting accuracy in practice.     

A decomposition-ensemble approach for tourism forecasting

With the frequent occurrence of irregular events in recent years, the tourism industry in some areas, such as Hong Kong, has suffered great volatility. To enhance the predictive accuracy of tourism demand forecasting, a decomposition-ensemble approach is developed based on the complete ensemble empirical mode decomposition with adaptive noise, data characteristic analysis, and the Elman's neural network model. Using Hong Kong tourism demand as an empirical case, this study firstly investigates how data characteristic analysis is used in a decomposition-ensemble approach. The empirical results show that the proposed model outperforms other models in both point and interval forecasts for different prediction horizons, indicating the effectiveness of the proposed approach for forecasting tourism demand, especially for time series with complexity.     

Forecasting tourism demand with denoised neural networks

The automated Neural Network Autoregressive (NNAR) algorithm from the forecast package in R generates sub-optimal forecasts when faced with seasonal tourism demand data. We propose denoising as a means of improving the accuracy of NNAR forecasts via an application into forecasting monthly tourism demand for ten European countries. Initially, we fit NNAR models on both raw and denoised (with Singular Spectrum Analysis) tourism demand series, generate forecasts and compare the results. Thereafter, the denoised NNAR forecasts are also compared with parametric and nonparametric benchmark forecasting models. Contrary to the deseasonalising hypothesis, we find statistically significant evidence which supports the denoising hypothesis for improving the accuracy of NNAR forecasts. Thus, it is noise and not seasonality which hinders NNAR forecasting capabilities.     

Modelling and forecasting the demand for Hong Kong tourism

The main objectives of this paper are to identify the factors which contribute to the demand for Hong Kong tourism with the aid of econometric models and to generate forecasts of international tourism arrivals to Hong Kong for the period 2001–2008. The general-to-specific modelling approach is followed to model and forecast the demand for Hong Kong tourism by residents from the 16 major origin countries/regions and the empirical results reveal that the most important factors that determine the demand for Hong Kong tourism are the costs of tourism in Hong Kong, the economic condition (measured by the income level) in the origin countries/regions, the costs of tourism in the competing destinations and the ‘word of mouth’ effect. The demand elasticities and forecasts of tourism arrivals obtained from the demand models form the basis of policy formulations for the tourism industry in Hong Kong.     

Impact of decomposition on time series bagging forecasting performance

Time series bagging has been deemed an effective way to improve unstable modelling procedures and subsequent forecasting accuracy. However, the literature has paid little attention to decomposition in time series bagging. This study investigates the impacts of various decomposition methods on bagging forecasting performance. Eight popular decomposition approaches are incorporated into the time series bagging procedure to improve unstable modelling procedures, and the resulting bagging methods' forecasting performance is evaluated. Using the world's top 20 inbound destinations as an empirical case, this study generates one-to eight-step-ahead tourism forecasts and compares them against benchmarks, including non-bagged and seasonal naïve models. For short-term forecasts, bagging constructed via seasonal extraction in autoregressive integrated moving average time series decomposition outperforms other methods. An autocorrelation test shows that efficient decomposition reduces variance in bagging forecasts.     

Forecasting tourism demand with multisource big data

Based on internet big data from multiple sources (i.e., the Baidu search engine and two online review platforms, Ctrip and Qunar), this study forecasts tourist arrivals to Mount Siguniang, China. Key findings of this empirical study indicate that (a) tourism demand forecasting based on internet big data from a search engine and online review platforms can significantly improve forecasting performance; (b) compared with tourism demand forecasting based on single-source data from a search engine, demand forecasting based on multisource big data from a search engine and online review platforms demonstrates better performance; and (c) compared with tourism demand forecasting based on online review data from a single platform, forecasting performance based on multiple platforms is significantly better.     

Forecasting tourism demand: Developing a general nesting spatiotemporal model

This study proposes a general nesting spatiotemporal (GNST) model in an effort to improve the accuracy of tourism demand forecasts. The proposed GNST model extends the general nesting spatial (GNS) model into a spatiotemporal form to account for the spatial and temporal effects of endogenous and exogenous variables as well as unobserved factors. As a general specification of spatiotemporal models, the proposed model provides high flexibility in modelling tourism demand. Based on a panel dataset containing quarterly inbound visitor arrivals to 26 European destinations, this empirical study demonstrates that the GNST model outperforms both its non-spatial counterparts and spatiotemporal benchmark models. This finding confirms that spatial and temporal exogenous interaction effects contribute to improved forecasting performance.     

Forecasting international tourism demand: a local spatiotemporal model

This study investigates whether tourism forecasting accuracy is improved by incorporating spatial dependence and spatial heterogeneity. One- to three-step-ahead forecasts of tourist arrivals were generated using global and local spatiotemporal autoregressive models for 37 European countries and the forecasting performance was compared with that of benchmark models including autoregressive moving average, exponential smoothing and Naïve 1 models. For all forecasting horizons, the two spatial models outperformed the non-spatial models. The superior forecasting performance of the local model suggests that the full reflection of spatial heterogeneity can improve the accuracy of tourism forecasting.     

Tourism demand forecasting with time series imaging: A deep learning model

To leverage computer vision technology to improve the accuracy of tourism demand forecasting, a model based on deep learning with time series imaging is proposed. The model consists of three parts: sequence image generation, image feature extraction, and model training. In the first part, the tourism demand data are encoded into images. In the second part, the convolution and pooling layers are used to extract features from the obtained images. In the final part, the extracted features are input into long short-term memory networks. Based on historical tourism demand data, the model for forecasting future tourism demand can be obtained. The performance of the proposed model is experimentally assessed through comparing against seven benchmark models.     

Econometric forecasts of tourism demand to 2005

A complete system of demand equations which was developed previously to generate forecasts of tourism imports and exports is modified to allow for destination-specific demand structures in the tourism export functions. The new model is shown to be considerably more realistic than the original one, and represents a major conceptual improvement. Furthermore, the modified complete system of demand equations yields more accurate outof-sample forecasts, across both varying time horizons and types of forecast. The new model is used to generate forecasts of tourism imports and exports for 18 countries and various major geographical areas, including the recently expanded European Union, for the period up to 2005 for different scenarios.     

Application of Cross-Impact Analysis in Tourism:

Qualitative approaches to forecasting generally are methods by which qualitative information form experts can be combined in order to make forecasts. Cross-impact analysis is such a qualitative forecasting method that is used form examining the impacts of potential future events upon each other. Given the complexity and interdependence which characterized tourism, cross-impact analysis seems to be a forecasting technique which is particularly relevant for tourism, and yet t appears to have been ignored in the tourism I the Azores, and illustrates how the technique may be used to assist with strategic planning.     

Forecasting accuracy evaluation of tourist arrivals

This paper evaluates the use of several parametric and nonparametric forecasting techniques for predicting tourism demand in selected European countries. We find that no single model can provide the best forecasts for any of the countries in the short-, medium- and long-run. The results, which are tested for statistical significance, enable forecasters to choose the most suitable model (from those evaluated here) based on the country and horizon for forecasting tourism demand. Should a single model be of interest, then, across all selected countries and horizons the Recurrent Singular Spectrum Analysis model is found to be the most efficient based on lowest overall forecasting error. Neural Networks and ARFIMA are found to be the worst performing models.     

Forecasting Chinese cruise tourism demand with big data: An optimized machine learning approach

After more than ten years of exponential development, the growth rate of cruise tourist in China is slowing down. There is increasingly financial risk of investing in homeports, cruise ships and promotional activities. Therefore, forecasting Chinese cruise tourism demand is a prerequisite for investment decision-making and planning. In order to enhance forecasting performance, a least squares support vector regression model with gravitational search algorithm (LSSVR-GSA) is proposed for forecasting cruise tourism demand with big data, which are search query data (SQD) from Baidu and economic indexes. In the proposed model, hyper-parameters of the LSSVR model are optimized with GSA. By comparing these models with various settings, we find that LSSVR-GSA with selected mobile keywords and economic indexes can achieve the highest forecasting performance. The results indicate the proposed framework of the methodology is effective and big data can be helpful predictors for forecasting Chinese cruise tourism demand.     

Visitor arrivals forecasts amid COVID-19: A perspective from the Asia and Pacific team

It is important to provide scientific assessments concerning the future of tourism under the uncertainty surrounding COVID-19. To this purpose, this paper presents a two-stage three-scenario forecast framework for inbound-tourism demand across 20 countries. The main findings are as follows: in the first-stage ex-post forecasts, the stacking models are more accurate and robust, especially when combining five single models. The second-stage ex-ante forecasts are based on three recovery scenarios: a mild case assuming a V-shaped recovery, a medium one with a V/U-shaped, and a severe one with an L-shaped. The forecast results show a wide range of recovery (10%–70%) in 2021 compared to 2019. This two-stage three-scenario framework contributes to the improvement in the accuracy and robustness of tourism demand forecasting.     

Predictivity of tourism demand data

As tourism researchers continue to search for solutions to determine the best possible forecasting performance, it is important to understand the maximum predictivity achieved by models, as well as how various data characteristics influence the maximum predictivity. Drawing on information theory, the predictivity of tourism demand data is quantitatively evaluated and beneficial for improving the performance of tourism demand forecasting. Empirical results from Hong Kong tourism demand data show that 1) the predictivity could largely help the researchers estimate the best possible forecasting performance and understand the influence of various data characteristics on the forecasting performance.; 2) the predictivity can be used to assess the short effect of external shock — such as SARS over tourism demand forecasting.     

Forecasting tourist arrivals using denoising and potential factors

Precise tourist demand forecasting is crucial owing to its relevance in tourism decision-making. This study proposes a novel model for tourist demand forecasting on the basis of denoising and potential factors. The denoising strategy is proposed to improve the tourist demand forecasting, and an effective and promising denoising method based on Hilbert–Huang transform is developed. Two case studies are conducted to verify the validity and predictability of the proposed model. Results indicate that denoising remarkably improves the forecasting accuracy, and the proposed denoising technique outperforms other approaches. Furthermore, the proposed model exhibits the most satisfactory forecasting performance among all benchmark models, as well as excellent scalability and stability.     

Tourism demand for Italy and the business cycle

This study provides a strategy for modelling the effect of the business cycle on tourism demand under the rationale that tourism cycles are heavily influenced by lagged effects of the overall business cycle. Using quarterly data on overnight stays in Italian hotels, both domestic and inbound between 1985 and 2004, we adopt a structural time series approach to evaluate two alternative models, the first with a latent cycle component (LCC) and the second based on specific economic explanatory variables (XCV). The two models are compared in terms of explanatory power, best-fit, residual diagnostics and forecasting ability. The results show similar performances. The policy implication is that the XCV model can be used for calibrating countercyclical interventions in tourism policy.     

A piecewise linear approach to modeling and forecasting demand for Macau tourism

In this paper, we construct and use a piecewise linear method to model and forecast, on a monthly basis, the demand for Macau tourism. Data over the period January 1991–December 2005 and a seasonally adjusted series for tourism demand are used. The study examines 4 forecasting horizons ranging from 6 to 24 months in advance. Mean absolute percentage errors and root mean square errors are adopted as criteria for evaluating the accuracy of the forecasting exercises. Finally, the forecasts of piecewise linear model are compared with those of autoregressive trend model, seasonal autoregressive integrated moving average and its arch-rival fractionally integrated autoregressive moving average models. The piecewise linear model is more accurate than the three benchmark models tested and the improvement is practically significant.     

Identifying a destination’s optimal tourist market mix: Does a superior portfolio model exist?

Extant tourism research has used various portfolio model types to determine optimal tourist market mixes which simultaneously maximize total tourist expenditure and minimise the instability of international inbound tourism demand. We analyse the three portfolio models that have been applied in the tourism literature: two varieties of a levels model (that use the level of tourist arrivals, or bed nights to quantify tourist activity) and a growth rates model (that deploys the growth in the level of tourist activity). Applying these models using per capita expenditure in four distinctively different destination countries (Australia, Greece, Japan, and USA), we demonstrate that the Levels Model 1 is superior to the Levels Model 2 and the Growth Rates Model. It produces solutions that provide noticeably higher tourist expenditure with less instability of international tourism demand than the status quo. Theoretical contributions and practical implications for tourism policy makers and destination marketers are discussed.