Search

The global tourism industry is struggling to recover from the COVID-19 pandemic. During the COVID-19 pandemic, daily tourism forecasting is more critical than ever before in supporting decisions and planning. Considering the changes in tourist psyche and behaviour caused by COVID-19, this study attempts to investigate whether the statistical modelling methods can work reliably under the new normal when travel restrictions are eased or lifted. To this end, we first compare the predictivity of daily tourism demand data before and during COVID-19, and observe heterogeneous impacts across different geographical scales. Then an improved multivariate & multiscale decomposition-ensemble framework is proposed to forecast daily tourism demand. The empirical study indicates the superiority and practicability of the proposed framework before and during COVID-19. Finally, we call for more research on the comparability of tourism demand forecasting. [ABSTRACT FROM AUTHOR]

Tourism managers and practitioners rely on accurate demand forecasting and well-informed management guidance. Given the pandemic's consequences on tourism, future analysis in the during-epidemic era is urgently needed. This study aims to achieve three goals combining the utilization of decomposition algorithms and deep learning models: 1) to investigate the changes in tourism demand according to seasonal fluctuations of various frequencies, 2) to improve modelling accuracy in tourism demand forecasting during non-crisis periods and in the peri-COVID-19 era, and 3) to analyze tourism demand evolution in the peri-Covid-19 era. The volume of domestic tourism in Hawaii is used as sample data for demonstration and validation. The empirical findings demonstrate that the framework provided in this study has excellent interpretability and forecasting accuracy, surpasses all benchmark models in terms of error calculation and statistical tests and can provide further insights into peri-Covid-19 demand analysis and management. [ABSTRACT FROM AUTHOR]

Daily tourism demand forecasting can provide important implications for the tourism industry. However, there exist limitations in applying traditional methods to forecast daily tourism demand because of the complex seasonal patterns and holiday effects. In this study, we introduce FB Prophet and apply it to the forecasting of daily tourism demand in the Jiuzhai Valley National Park and Macao from Mainland China. The decomposition result of FB Prophet shows its ability to handle the influence of seasonal patterns and holiday effects. The forecasting results show that considering seasonal patterns, holiday effects, and other predictors can significantly improve the forecasting performance, and FB Prophet outperforms other methods. [ABSTRACT FROM AUTHOR]

Existing research has shown that combination can effectively improve tourism forecasting accuracy compared with single model. However, the model uncertainty and structural instability in combination for out-of-sample tourism forecasting may influence the forecasting performance. This paper proposes a novel forecast combination approach based on time-varying jackknife model averaging (TVJMA), which can more efficiently handle structural changes and nonstationary trends in tourism data. Using Hong Kong tourism demand from five major tourism source regions as an empirical study, we investigate whether our proposed nonparametric TVJMA-based approach can improve tourism forecasting accuracy further. Empirical results show that the proposed TVJMA-based approach outperforms other competitors including single model and three combination methods in most cases. Findings indicate the outstanding performance of our method is robust to various forecasting horizons and different estimation periods. [ABSTRACT FROM AUTHOR]

The availability of tourism-related big data increases the potential to improve the accuracy of tourism demand forecasting but presents significant challenges for forecasting, including curse of dimensionality and high model complexity. A novel bagging-based multivariate ensemble deep learning approach integrating stacked autoencoder and kernel-based extreme learning machine (B-SAKE) is proposed to address these challenges in this study. By using historical tourist arrival data, economic variable data, and search intensity index (SII) data, we forecast tourist arrivals in Beijing from four countries. The consistent results of multiple schemes suggest that our proposed B-SAKE approach outperforms the benchmark models in terms of level accuracy, directional accuracy, and even statistical significance. Both bagging and stacked autoencoder can effectively alleviate the challenges brought by tourism big data and improve the forecasting performance of the models. The ensemble deep learning model we propose contributes to tourism demand forecasting literature and benefits relevant government officials and tourism practitioners. [ABSTRACT FROM AUTHOR]

Decomposition methods are extensively used for processing the complex patterns of tourism demand data. Given tourism demand data's intrinsic complexity, it is critical to theoretically understand how different decomposition methods provide solutions. However, a comprehensive comparison of decomposition methods in tourism demand forecasting is still lacking. Hence, this study systematically investigates the forecasting performance of decomposition methods in tourism demand. Nine popular decomposition methods and six forecasting methods are employed, and their forecasting performance is compared. With Hong Kong visitor arrivals from eight major sources as a sample, three main conclusions are obtained from empirical results. First, all the decomposition methods generally outperform benchmark at all horizons, in both the level and directional forecasting. Second, decomposition methods can be divided into four categories based on forecasting accuracy. Finally, variational mode decomposition method is consistently superior to other eight decomposition methods and can provide the best forecasts in all cases. [ABSTRACT FROM AUTHOR]

Search query data (SQD) can be helpful in predicting tourism demand by generating web search indexes. However, valuable nonlinear information in SQD may be neglected by researchers. To effectively capture the nonlinear information, we used kernel principal component analysis (KPCA) to extract web search indexes from SQD. Then, several models with KPCA-based web search indexes were developed for tourism demand forecasting. An empirical study was conducted with collected SQD and real data of tourist arrivals at Hong Kong. The results suggest that models with KPCA-based web search indexes are more accurate than other models because of the nonlinear data processing ability of the KPCA and demonstrate that KPCA-based web search indexes can be excellent predictors for tourism demand forecasting. [ABSTRACT FROM AUTHOR]

The coronavirus disease (COVID-19) pandemic has already caused enormous damage to the global economy and various industries worldwide, especially the tourism industry. In the post-pandemic era, accurate tourism demand recovery forecasting is a vital requirement for a thriving tourism industry. Therefore, this study mainly focuses on forecasting tourist arrivals from mainland China to Hong Kong. A new direction in tourism demand recovery forecasting employs multi-source heterogeneous data comprising economy-related variables, search query data, and online news data to motivate the tourism destination forecasting system. The experimental results confirm that incorporating multi-source heterogeneous data can substantially strengthen the forecasting accuracy. Specifically, mixed data sampling (MIDAS) models with different data frequencies outperformed the benchmark models. [ABSTRACT FROM AUTHOR]

Abstract Previous studies have shown that online data, such as search engine queries, is a new source of data that can be used to forecast tourism demand. In this study, we propose a forecasting framework that uses machine learning and internet search indexes to forecast tourist arrivals for popular destinations in China and compared its forecasting performance to the search results generated by Google and Baidu, respectively. This study verifies the Granger causality and co-integration relationship between internet search index and tourist arrivals of Beijing. Our experimental results suggest that compared with benchmark models, the proposed kernel extreme learning machine (KELM) models, which integrate tourist volume series with Baidu Index and Google Index, can improve the forecasting performance significantly in terms of both forecasting accuracy and robustness analysis. Highlights • We propose kernel extreme learning machine with different kernel functions to accurately forecast tourist arrivals. • Co-integration and granger causality relationship between search engine query data and tourist arrivals are verified. • Baidu index and Google trends data are simultaneously regarded as exogenous variables to forecast Beijing tourist arrivals. • Our proposed approach significantly improves the forecasting performance better than other benchmarks used in this study. [ABSTRACT FROM AUTHOR]

Precise forecasts of railway container demand is crucial for railway infrastructure construction and daily operations management. Nevertheless, temporal variations in railway container traffic are subject to a diversity of factors, and such dynamics are therefore extremely difficult to capture in forecasts. Decomposing the temporal data into a number of subseries with distinct qualities is a prevalent technique for reducing the complexity of data. Yet, a few of sub-series resulted from such decomposition tend to remain irregular and unstable. In this study, a novel secondary decomposition–ensemble method is proposed for railroad container volume forecasting. Specifically, CEEDMAN and VMD are employed successively to decompose the time series into various components. The sub-series are processed by different models, according to their distinct features, to yield their respective independent predictions, which are then ensembled to give the final railway container traffic forecast. Additionally, we incorporate the DEEPAR model for interval predictions, providing valuable uncertainty information that assists decision-makers in preparing for potential challenges. Our method demonstrates superior accuracy and stability compared to other benchmark models, even in the presence of disruptive events such as the big epidemic. We further validate the applicability and robustness of the proposed method by incorporating an additional dataset with a finer-time data. Our findings contribute to the field of railway container demand forecasting and offer practical insights for decision-makers in the railway industry. • Minimized time-series data complexity via twice decomposition. • Sub-sequences generated by the decomposition exhibit distinctive patterns, processed individually by identifying the most suitable prediction model. • Provided point & interval projections for railway container volume, with interval indicating uncertainty. • Model exhibited exceptional accuracy, robustness, resistance to noise, and stable performance during big events. [ABSTRACT FROM AUTHOR]

Abstract: Air transport demand forecasting is receiving increasing attention, especially because of intrinsic difficulties and practical applications. Total passengers are used as a proxy for air transport demand. However, the air passenger time series usually has a complex behavior due to their irregularity, high volatility and seasonality. This paper proposes a new hybrid approach, combining singular spectrum analysis (SSA), adaptive-network-based fuzzy inference system (ANFIS) and improved particle swarm optimization (IPSO), for short-term air passenger traffic prediction. The SSA is used for identifying and extracting the trend and seasonality of air transport demand and the artificial intelligence technologies, including ANFIS and IPSO, are utilized to deal with the irregularity and volatility of the demand. The HK air passenger data are collected to establish and validate the forecasting model. Empirical results clearly points to the enormous potential that the proposed approach possesses in air transport demand forecasting and can be considered as a viable alternative. [Copyright &y& Elsevier]

• A two stage-based feature selection approach for forecasting is proposed. • The approach relies on machine learning model combined with feature selection method. • With different forecasting horizons and datasets, results are predominant. • The study confirms the value of exploring the more effective keywords subset. • The results provide some implications for forecasting using multisource data. Search engine data have been widely used and shown to be useful in tourism demand forecasting. However, considering of the vast amounts of search keywords, how to better capture the tourists' attention and explore the most predictive keyword combination remain unsolved. In this study, a two-stage feature selection-based methodology is proposed to address this question. Specifically, i.e., single feature selection method comparison for selecting a relative effective way to reduce the data dimension and ensure the quality of the initial subset, genetic algorithm in the second stage for obtaining feature subset better suitable for forecasting model with stronger predictive power. Experimental results indicate that the two-stage feature selection method outperforms all the considered benchmarks. [ABSTRACT FROM AUTHOR]

• Multi-step ahead demand forecasting is an important and cutting-edge research topic. • An effective forecasting approach incorporating privileged information is proposed. • The new paradigm (termed learning using privileged information (LUPI)) is imployed. • LUPI can efficiently grasp and model nonlinear characteristics of tourism data. • The empirical results verify its superiority and robustness with different samples. Accurate tourism demand forecasting can provide effective guidance for government management and tourism planning. Specifically, multi-step ahead tourism demand forecasting is of great relevance in guiding managers to develop strategies and assisting operators in business planning and is therefore of great interest to researchers and practitioners. However, it is usually difficult to obtain satisfactory accuracy in views of sophisticated data characteristics. This study proposes an improved machine learning paradigm, introducing valuable additional information into the training phase, which is not available for forecasting in the testing phase. Taking Hawaii (Daily and Weekly) and Macau tourist arrivals as the samples, the empirical evidence indicates that the proposed approach can significantly enhance the multi-step ahead forecasting performance from the view of both error calculation and statistical test. In particular, the paradigm's robustness is also demonstrated in the comparison of sliding window predictions with models of variants. [ABSTRACT FROM AUTHOR]

Current research highlights the efficacy of decomposition and ensemble algorithms in enhancing forecasting accuracy; however, the investigation of robustness associated with decomposed components within these algorithms remains notably scarce in existing literature. To address this gap, we introduce a novel tourism demand forecasting framework, underpinned by a sophisticated decomposition algorithm. Our approach initially decomposes the original data into multiple sub-series and subsequently selects forecasting models based on their respective data attributes. We evaluated the proposed framework by forecasting monthly tourist arrivals in Hong Kong from six countries. The superiority of the novel decomposition method was further substantiated through comparisons with alternative decomposition techniques in both single-step and multi-step ahead forecasting contexts. The results indicate that the proposed forecasting framework consistently outperforms baseline models in terms of forecasting accuracy across all tourism demand forecasting scenarios, with the innovative decomposition method exhibiting exceptional performance. [ABSTRACT FROM AUTHOR]

• Strategies of previous tourism literature are summarized. • A meticulous review and comparison of five strategies is conducted. • The performance of five forecasting strategies is explored and compared. • The DIRMO strategy is most practical among all these strategies. • Empirical findings provide a reference paradigm for future research. Previous researches have proposed five strategies to deal with complex multi-step ahead forecasting tasks. However, these strategies have not received much attention in the field of tourism research and the performance of them is still unknown. Accordingly, we summarize the strategies used in multi-step ahead tourism demand forecasting articles and produce a comparative analysis of five strategies. By employing nine tourist arrival time series in Hong Kong, the study explores the performance of different strategy-driven forecasting approaches in tourism demand. The empirical results show that Direct and DIRMO (s = 2) strategies perform better than other strategies in terms of forecasting accuracy, but there is no significant difference between them, excepting the latter is lower than the former in computational complexity. [ABSTRACT FROM AUTHOR]

Accurate electricity consumption forecasting plays a significant role in power production and supply and power dispatching. Thus, a new hybrid model combing a grey model with fractional order accumulation, called FGM (1, 1), with seasonal factors, sine cosine algorithm (SCA), and an error correction strategy is proposed in this research. To accurately predict the seasonal fluctuations, seasonal factors are used in this model; Then, with the aim of improving the prediction performance, a SFGM (1, 1) model optimized by SCA rather than least square method, namely SCA-SFGM (1, 1), is establish to forecast electricity consumption; Moreover, considering forecasting error sequence may contain useful information, an error correction strategy is introduced to model forecasting error time series to adjust the preliminary forecasts of SCA-SFGM (1, 1). Fourth, four comparison models, three measurement criteria and a statistical hypothesis testing method using monthly residential electricity consumption dataset from 2015 to 2020 are designed to verify the prediction performance of models; Lastly, experimental results show that the mean absolute percentage error (MAPE) of the proposed model is 4.1698%, which is much lower than 14.5642%, 6.5108%, 5.9472%, 5.7060% and 4.9219% of GM (1, 1), SARIMA, SGM (1, 1), SFGM (1, 1) and SCA-SFGM (1, 1) models, respectively, showing that the proposed model can not only effectively capture seasonal fluctuations, it also adds an operational candidate forecasting benchmark model in electricity markets. [Display omitted] • Proposed a model using an optimized SFGM and an error correction strategy. • The proposed model can reveals the changing trends and seasonal fluctuations. • Three criteria, five benchmark models and a hypothesis testing are constructed. • The proposed model is proved to have higher accuracy than comparison models. [ABSTRACT FROM AUTHOR]

• Site dependency and distribution heterogeneity of wind speed data is considered. • Generalized Gaussian process models with positive domain likelihoods are proposed. • A statistical model is selected to extract the linear feature from the wind speed. • A pseudo-likelihood with cross validation is applied to improve models' robustness. • Empirical results statistically verify the performance of our proposed models. As the proportion of the total installed wind capacity continues to increase, precise probabilistic and deterministic wind speed forecasting becoming increasingly significant for wind turbine stable function and operation management. It is worth noting that except for the intermittent nature of wind speed itself, the site dependence of wind energy and the heterogeneous nature of the distribution at different locations can also affect the results of wind speed forecasting. Considering this has often been overlooked in previous studies, this study attempts to introduce characteristic information of wind from wind farms into wind speed probabilistic forecasting. For the sake of acquiring a more rational descriptive statistical interpretation of wind speed, a variety of probability density functions for actual wind speed data are evaluated in this study. And then, Generalized Gaussian Process models are constructed via observation likelihood functions which are determined by obtained probability density functions. Beyond that, to enhance the robustness against model misspecification, a pseudo-likelihood for the Leave-One-Out cross-validation methodology is proposed for approximating the posterior in the Bayesian inference stage. Wind speed forecasts on 6-hour-ahead horizons from two wind farms in China illustrated that the appropriate mining of wind prior statistical characteristic information can contribute to improving the precision of probabilistic wind speed forecasting. [ABSTRACT FROM AUTHOR]