Ensemble Forecasting Approach Research Articles

A potential threat to amphibians in the European Natura 2000 network: Forecasting the distribution of the American bullfrog Lithobates catesbeianus

Freshwater biodiversity is declining at an accelerated pace. Climate change and associated global warming and changes in precipitation patterns, combined with the expansion of generalist -invasive species are two of the main threats. Niche-based models (NBMs) are becoming inevitable tools in invasive species risk assessment and in conservation decision-making. Lithobates catesbeianus is an invasive species globally known for its adverse ecological impacts on native amphibians and biodiversity. To assess species current and future climatic suitable areas at the global and European scales we used an ensemble forecasting approach. We considered six climatic variables, three timeframes (current, 2050, and 2070), and two CO2 emission scenarios. Temperature seasonality, minimum temperature of the coldest month, maximum temperature of the warmest month, and precipitation in the driest month were the most important variables predicting bullfrog occurrence. Globally currently 3.8% of land area is suitable for bullfrog and an increase of up to 5.2% in 2070 is expected. Increase in suitable areas is expected at higher latitudes, especially in North America and central Europe. Currently, 3.45% of total Natura 2000 area is suitable, and a predicted range gain of up to 355.93% (12.28%) is expected in the highest concentration scenarios predictions. This can indicate that the 64 native amphibian species present in the Natura 2000 network could be at increased risk. The choice of Natura 2000 for a geographic detailed analysis of the possible effects on native amphibians is due to its importance for habitats and wildlife conservation. Identification of its invasion-susceptible areas will allow resource and management practices optimization.

An Accurate Ensemble Forecasting Approach for Highly Dynamic Cloud Workload With VMD and R-Transformer

To efficiently manage the cloud resources, improve the quality of service, and avoid the violations of Service-Level Agreement (SLA) agreements, it is very important to make accurate forecast for cloud workload. Prior works concerning cloud workload forecasting are mainly designed based on Recurrent Neural Networks (RNN). However, when it comes to a highly-dynamic cloud workload scenario where resource utilization changes faster and more frequently, these RNN-based methods are not effective in obtaining the linear and non-linear relationships and cannot give accurate forecast, because classic RNN has the problem of vanishing gradient. To address this issue, we propose an Ensemble Forecasting Approach (EFA) for highly-dynamic cloud workload by applying Variational Mode Decomposition (VMD) and R-Transformer. Specifically, to decrease the non-stationarity and high randomness of highly-dynamic cloud workload sequences, we decompose the workload into multiple Intrinsic Mode Functions (IMFs) by VMD. The IMFs are then imported into our ensemble forecasting module based on R-Transformer and Autoregressive model, in order to capture long-term dependencies and local non-linear relationship of workload sequences. The effectiveness and adaptability of proposed EFA is verified on real-world workload from Google and Alibaba cluster traces. Moreover, the performance evaluation results show that the EFA performs higher forecasting accuracy than prior related works over various forecasting time lengths for highly-dynamic cloud workload.

Accuracy of real-time multi-model ensemble forecasts for seasonal influenza in the U.S.

Seasonal influenza results in substantial annual morbidity and mortality in the United States and worldwide. Accurate forecasts of key features of influenza epidemics, such as the timing and severity of the peak incidence in a given season, can inform public health response to outbreaks. As part of ongoing efforts to incorporate data and advanced analytical methods into public health decision-making, the United States Centers for Disease Control and Prevention (CDC) has organized seasonal influenza forecasting challenges since the 2013/2014 season. In the 2017/2018 season, 22 teams participated. A subset of four teams created a research consortium called the FluSight Network in early 2017. During the 2017/2018 season they worked together to produce a collaborative multi-model ensemble that combined 21 separate component models into a single model using a machine learning technique called stacking. This approach creates a weighted average of predictive densities where the weight for each component is determined by maximizing overall ensemble accuracy over past seasons. In the 2017/2018 influenza season, one of the largest seasonal outbreaks in the last 15 years, this multi-model ensemble performed better on average than all individual component models and placed second overall in the CDC challenge. It also outperformed the baseline multi-model ensemble created by the CDC that took a simple average of all models submitted to the forecasting challenge. This project shows that collaborative efforts between research teams to develop ensemble forecasting approaches can bring measurable improvements in forecast accuracy and important reductions in the variability of performance from year to year. Efforts such as this, that emphasize real-time testing and evaluation of forecasting models and facilitate the close collaboration between public health officials and modeling researchers, are essential to improving our understanding of how best to use forecasts to improve public health response to seasonal and emerging epidemic threats.

Shifting ranges of eleven invasive alien plants in China in the face of climate change

Identifying the factors predicting the suitable habitats of invasive alien plants (IAPs) and how these suitable habitats are affected by climate change are important for taking measures to limit the spread of IAPs in China. In this study, an ensemble forecasting approach is used to accurately evaluate the relationships between IAPs distribution and the environmental variables, and to simulate the range shifts of the eleven invasive species under different climate scenarios. All of the AUC and TSS values from ensemble forecasting models are above 0.9, indicating its perfect ability to distinguish actual suitable and unsuitable habitat. Our models show that annual mean temperature and temperature seasonality have strong correlations with alien weeds distribution. The results indicate that all the 11 IAPs ranges will expand northward, which may be due to global climate warming. The prediction shows that a more aggressive climate scenario will be more conducive to the northward expansion of IAPs. The results of this study can provide a new insight for policy makers to monitor and control IAPs.

Modelling potential distribution of fluvial fish species for expanding conservation knowledge: Case study of the genus Barbus in Iran

Species inhabiting fresh waters are severely influenced by anthropogenic factors. Effective management and conservation plans require high accurate and reliable species distribution forecasts. Here, we modelled potential distribution of the genus Barbus in Iran, based on environmental variables using Species Distribution Models (SDMs). Six environmental predictors (i.e. slope, bankfull width, elevation, mean air temperature, range of air temperature and annual precipitation) were applied for modelling. The models were selected among different technique (GLM, GAM, CTA, SRE, GBM, RF, MARS, and FDA) which their results were summarized through ensemble forecasting approaches. According to the TSS (True Skill Statistic), the accuracy of the implemented models was greater than 0.8. The results showed that the projected distributions not only were observed in the same recorded basins but also in the new basins. Presented results deepen the conservation knowledge in Iran and act as a guidance for management decisions aimed at legal identification of critical habitats for species as well as informing them for translocation of threatened or captive-bred populations.

Improving spatial accuracy of urban growth simulation models using ensemble forecasting approaches

This paper aims to improve the spatial accuracy of urban growth simulation models and clarify any associated uncertainties. Artificial Neural Networks (ANNs), Random Forest (RF), and Logistic Regression (LR) were implemented to simulate urban growth in the megacity of Tehran, Iran, as a case study. Model calibration was performed using data between 1985 and 1999 whereas the data between 1999 and 2014 was used for model validation. First of all, Transition Index Maps (TIMs) were computed by means of each model to assess the potential of urban growth for each cell. Using the standard deviation, consensus between the TIMs was evaluated. Because the TIMs of the individual models manifested discrepancies, they were combined using a number of ensemble forecasting approaches including median, mathematical average, principle component analysis, and weighted area under the total operating characteristic. The individual and combined TIMs were then put into Cellular Automata (CA) to simulate the future pattern of urban growth in Tehran. The results were evaluated in two stages. At first, the TIMs were evaluated by means of Total Operating Characteristics (TOC), and then a set of statistical indices was used to evaluate the spatial accuracy of the simulated urban growth maps. The best result was obtained by median ensemble forecasting, whereas the LR model showed the lowest level of accuracy. In similar studies, it is recommended to implement and compare different ensemble methods when integrating individual models.

Modelling climate change effects on Zagros forests in Iran using individual and ensemble forecasting approaches

It is believed that climate change will cause the extinction of many species in the near future. In this study, we assessed the impact of climate change on the climatic suitability of the Persian oak in Zagros forests in southwest Iran, by simulating their conditions under four climate change scenarios in the 2030s, 2050s, 2070s, and 2080s. Additionally, we evaluated the predictive performance of different modelling algorithms by projecting the geographic distribution of Persian oak, using a block cross-validation technique. According to the results, the Persian oak shows a stronger response to temperature, particularly the maximum temperature of the warmest month, rather than precipitation variables. This indicates that temperature has a powerful control over the geographic distribution of the Persian oak. Based on a comparison of the Persian oak’s current climatic suitability and future projections, regardless of the chosen climatic scenarios, there will be an upward shift in its climatic suitability. However, an upward shift under the pessimistic scenarios was greater than the optimistic ones. The results also indicate that an ensemble of all models had a higher accuracy than single models. Despite the agreement between current climate condition predictions (mean correlation of 0.94), the projection of different algorithms for future periods is highly variable (mean correlation of 0.71). Thus, the ensemble approach was used to reduce the uncertainty of modelling, favouring the consensus of all models for future projection.

Effective sparse adaboost method with ESN and FOA for industrial electricity consumption forecasting in China

Accurate electricity consumption forecasting is a challenging task for its unstable behavior and influence mechanism based on multiple factors. In this study, a neural network ensemble approach is designed to solve this problem. In the proposed method, a novel sparse adaboost (adaboostsp) is designed as the ensemble framework to enhance the generalization ability and reduce ensemble cost, and echo state network (ESN) is adopted to build the nonlinear relationships between electricity demand and multiple factors. An improved fruit fly optimization algorithm (FOA) helps selecting input variables considering their time lag effects. Two industrial electricity consumption (IEC) forecasting applications in China are investigated to verify the effectiveness of proposed ensemble forecasting approach. Numerical results indicate that adaboostsp-ESN with FOA can better predict the future IEC than various benchmark methods. Compared with existing boosting ensemble approaches, the proposed adaboostsp is more efficient and can save considerable computation cost. Impacts of selected variables are further examined and results show many industrial indexes have significant time lag effects on IEC. Based on the proposed techniques, future IEC demand in Hubei Province is estimated and analyzed. Application studies demonstrate the proposed hybrid ensemble approach is a practical choice for mid-term IEC adjustment and projection.

Novel forecasting approaches using combination of machine learning and statistical models for flood susceptibility mapping

In this research, eight individual machine learning and statistical models are implemented and compared, and based on their results, seven ensemble models for flood susceptibility assessment are introduced. The individual models included artificial neural networks, classification and regression trees, flexible discriminant analysis, generalized linear model, generalized additive model, boosted regression trees, multivariate adaptive regression splines, and maximum entropy, and the ensemble models were Ensemble Model committee averaging (EMca), Ensemble Model confidence interval Inferior (EMciInf), Ensemble Model confidence interval Superior (EMciSup), Ensemble Model to estimate the coefficient of variation (EMcv), Ensemble Model to estimate the mean (EMmean), Ensemble Model to estimate the median (EMmedian), and Ensemble Model based on weighted mean (EMwmean). The data set covered 201 flood events in the Haraz watershed (Mazandaran province in Iran) and 10,000 randomly selected non-occurrence points. Among the individual models, the Area Under the Receiver Operating Characteristic (AUROC), which showed the highest value, belonged to boosted regression trees (0.975) and the lowest value was recorded for generalized linear model (0.642). On the other hand, the proposed EMmedian resulted in the highest accuracy (0.976) among all models. In spite of the outstanding performance of some models, nevertheless, variability among the prediction of individual models was considerable. Therefore, to reduce uncertainty, creating more generalizable, more stable, and less sensitive models, ensemble forecasting approaches and in particular the EMmedian is recommended for flood susceptibility assessment.

Using a multi-model ensemble forecasting approach to identify key marine protected areas for seabirds in the Portuguese coast

Marine Protected Areas (MPAs) have been established across all marine environments, though their coherence and effectiveness in protecting umbrella species remains unclear. We used a multi-model ensemble forecasting approach, on 8 years of at-sea censuses of 30 seabird species to identify candidate MPAs in the Portuguese coast, prioritizing important areas for their conservation based on their occurrence and distribution. We overlapped the outputs generated by the Ensemble Ecological Niche Models (EENMs) with layers representing important environmental stressors (fishing intensity, ship density and oil pollution risk), and calculated loss in conservation value using them as cost layers. Three key marine areas were identified along the Portuguese coast: For breeders, there was a key marine area encompassing the Tagus and Sado estuaries and Berlengas archipelago; for non-breeders and migratory species two important areas were identified in the Northern and Southern coast. The key marine area identified in the Northern coast is characterized by high productivity and biodiversity, and can be affected by oil pollution from the refineries and the intensive ship traffic in this area. Also, the area identified in the Southern coast of Portugal for migratory seabirds overlaps extensively with areas of high fishing activity. Our results show that the Important Bird and Biodiversity Areas (IBAs) established along the Portuguese coast protect more than a third of the areas that we prioritized for breeding species and the official MPAs near 65% of the same areas. In contrast, current IBAs and national legislation protect less than 4% of the coastal areas that we prioritize for non-breeding species in this study. Our study, combining multi-species distribution with environmental constraints induced by human activities, allowed us to assess the coherence of the Portuguese marine planning and identify candidate areas to join the Portuguese network of marine protected areas. Our results, employing data from annual at-sea surveys together with the human stressors known to affect the Portuguese coast, proved to be an extremely useful strategy to identify spatial conservation areas along the Portuguese coast as well as to access the adequacy and consistency of those areas. Despite the constraints of this demanding approach, we are confident that our study provides a reliable strategy to inform marine conservation efforts and management planning in similar coastal environments elsewhere, characterized by strong coastal upwelling movements.

Land use change uncertainty impacts on streamflow and sediment projections in areas undergoing rapid development: A case study in the Mekong Basin

AbstractQuantitative understanding of potential changes in streamflow and sediment load is complicated by uncertainty related to land use change projections, which is characterized by a high uncertainty in terms of demand (quantity) and location of changes (spatial distribution). We simulate the Sesan, Srepok, and Sekong Rivers (3S), the most important tributaries of the lower Mekong River, with the Soil and Water Assessment Tool (SWAT) to investigate the implications of conversion of forest to agricultural lands. Multiple land use transitions in the 3S basin are projected using the Land Change Modeler. The uncertainty in land use projection was addressed using an ensemble forecasting approach for 2060, combining (a) three land demand scenarios, (b) two transition potential modeling approaches (i.e., approach to create maps of the likelihood for areas to transition from one land use type to another), and (c) retaining or not protected areas. Land demand leads to the greatest uncertainty in land use change projections. Transition potential modeling approaches do not make much difference in the total change, but can result in spatial variations of change. Retaining protected areas can contribute significantly to uncertainty in land use change projections. Decrease in annual streamflow of the 3S basin varied from 3% to 21%, and changes in annual sediment outflux from the basin ranged from −8% to 249% for simulated scenarios. Land use demand uncertainty results in the highest streamflow and sediment load changes and can thus have major consequences for water and sediment management strategies in areas undergoing rapid development.

Ensemble forecasting of Corbicula fluminea worldwide distribution: Projections of the impact of climate change

Abstract Global biodiversity is at risk owing to climate change, and freshwater ecosystems are expected to suffer the most. In recent years niche‐based models (NBMs) have been used to predict species distribution and are an important tool for conservation and management of aquatic ecosystems. In this work, the current and future climatic suitability areas of the invasive species Corbicula fluminea, which has known adverse ecological and economic impacts, were investigated. The species distribution modelling was based on nine algorithms in BIOMOD2, summarized in an ensemble forecasting approach. To model the species distribution, eight climatic parameters related to temperature and precipitation variables were considered. Three time frames (current, 2050 and 2070) were modelled using four increasing CO2 emission scenarios. The performance of individual models was excellent according to the area under the receiver operating characteristic curve (AUC) and good to excellent according to true skill statistics (TSS). Annual mean temperature, minimum temperature of the coldest month and mean temperature of the coldest quarter were the most important variables predicting C. fluminea occurrence. Of the total continental area, 6.6% was predicted to be suitable for C. fluminea in current conditions. In the future, suitable area will increase from the current value of 6.6% to values from 9.4% to 12.6%, according to the 2050 projections and up to 12.7% in 2070 in high emission scenarios. Overall, the results indicate that climate change will favour the expansion of C. fluminea into new river basins, especially at higher latitudes, and that future climatic scenarios may double the suitable area for Corbicula fluminea.

Data-Driven Frequency-Based Airline Profit Maximization

Although numerous traditional models predict market share and demand along airline routes, the prediction of existing models is not precise enough, and to the best of our knowledge, there is no use of data mining--based forecasting techniques for improving airline profitability. We propose the maximizing airline profits (MAP) architecture designed to help airlines and make two key contributions in airline market share and route demand prediction and prediction-based airline profit optimization. Compared to past methods used to forecast market share and demand along airline routes, we introduce a novel ensemble forecasting (MAP-EF) approach considering two new classes of features: (i) features derived from clusters of similar routes and (ii) features based on equilibrium pricing. We show that MAP-EF achieves much better Pearson correlation coefficients (greater than 0.95 vs. 0.82 for market share, 0.98 vs. 0.77 for demand) and R 2 -values compared to three state-of-the-art works for forecasting market share and demand while showing much lower variance. Using the results of MAP-EF, we develop MAP--bilevel branch and bound (MAP-BBB) and MAP-greedy (MAP-G) algorithms to optimally allocate flight frequencies over multiple routes to maximize an airline’s profit. We also study two extensions of the profit maximization problem considering frequency constraints and long-term profits. Furthermore, we develop algorithms for computing Nash equilibrium frequencies when there are multiple strategic airlines. Experimental results show that airlines can increase profits by a significant margin. All experiments were conducted with data aggregated from four sources: the U.S. Bureau of Transportation Statistics (BTS), the U.S. Bureau of Economic Analysis (BEA), the National Transportation Safety Board (NTSB), and the U.S. Census Bureau (CB).

Ensemble Forecasting for Complex Time Series Using Sparse Representation and Neural Networks

AbstractBased on the concept of ‘decomposition and ensemble’, a novel ensemble forecasting approach is proposed for complex time series by coupling sparse representation (SR) and feedforward neural network (FNN), i.e. the SR‐based FNN approach. Three main steps are involved: data decomposition via SR, individual forecasting via FNN and ensemble forecasting via a simple addition method. In particular, to capture various coexisting hidden factors, the effective decomposition tool of SR with its unique virtues of flexibility and generalization is introduced to formulate an overcomplete dictionary covering diverse bases, e.g. exponential basis for main trend, Fourier basis for cyclical (and seasonal) features and wavelet basis for transient actions, different from other techniques with a single basis. Using crude oil price (a typical complex time series) as sample data, the empirical study statistically confirms the superiority of the SR‐based FNN method over some other popular forecasting models and similar ensemble models (with other decomposition tools). Copyright © 2016 John Wiley & Sons, Ltd.

Topic discovery and future trend forecasting for texts

Finding topics from a collection of documents, such as research publications, patents, and technical reports, is helpful for summarizing large scale text collections and the world wide web. It can also help forecast topic trends in the future. This can be beneficial for many applications, such as modeling the evolution of the direction of research and forecasting future trends of the IT industry. In this paper, we propose using association analysis and ensemble forecasting to automatically discover topics from a set of text documents and forecast their evolving trend in a near future. In order to discover meaningful topics, we collect publications from a particular research area, data mining and machine learning, as our data domain. An association analysis process is applied to the collected data to first identify a set of topics, followed by a temporal correlation analysis to help discover correlations between topics, and identify a network of topics and communities. After that, an ensemble forecasting approach is proposed to predict the popularity of research topics in the future. Our experiments and validations on data with 9 years of publication records validate the effectiveness of the proposed design.

Neotropical wild cats susceptibility to climate change

Ongoing climate change and the human role as dominant cause behind it are undeniable and already affecting living systems around the globe. Nonetheless, the likely consequences of climate change to Neotropical biodiversity are still poorly understood. We used species distribution modeling to evaluate the likely effects of climate change to the seven species of wild cats that are endemic to the Neotropics. We gathered (and provide) 424 species occurrence records from museum collections and the literature. We run the analysis on the ModEco software, using four modeling algorithms and projected models into 2050 using data from International Panel on Climate Change's last Assessment Report, under a business-as-usual emission scenario (RCP 8.5), according to five Global Circulation Models. We used an ensemble-forecasting approach to reach a consensus scenario, including only models with AUC > 0.70 for the present climate dataset. We created ensembles using the majority rule. After this procedure, we ended with two final suitability models per species, one for the present and another for the future. Model performance varied among species and was related to species' climatic suitability area (the smaller the area, the greater the model performance), and species with the smaller ranges were predicted to lose the highest percentage of their current distribution under climate change. The projections under climate change points to important contraction of climatically suitable areas for all Neotropical felids except for L. geoffoyii . The remaining species show, in average, a 43.5% contraction of suitable areas, with L. jacobitus and L. guigna showing more than 50% contraction. Both are already threatened under IUCN and Leopardus jacobitus , found only in the higher elevations of the Andes, is of special concern because highland species are particularly susceptible to a warming climate.

FishMed: traits, phylogeny, current and projected species distribution of Mediterranean fishes, and environmental data

The FishMed database provides traits, phylogeny, current and projected species distribution of Mediterranean fishes, and associated sea surface temperature (SST) from the regional oceanic model NEMOMED8. Data for the current geographical distributions of 635 Mediterranean fish species were compiled from a published expert knowledge atlas of fishes of the northern Atlantic and the Mediterranean (FNAM) edited between 1984 and 1986 and from an updated exotic fish species list. Two future sets of projected species distributions were obtained for the middle and end of the 21st century by using an ensemble forecasting approach for 288 coastal Mediterranean fish species based on SST according to the IPPC/SRES A2 scenario implemented with the Mediterranean climatic model NEMOMED8. The functional part of the database encompasses 12 biological and ecological traits (maximal and common lengths, vertical distribution, habitat, migration type, mode of reproduction, sex shift, semelparity, diet type (larvae and adults), social behavior, species origin, and depth) for the 635 fish species. To build the phylogeny we inferred the timing and geographic origins of Mediterranean teleost species diversity using nucleotide sequences collected from GenBank including 62% of Mediterranean teleost species plus nine outgroups. Maximum likelihood Bayesian phylogenetic and dating analyses were calibrated using 20 fossil species. An additional 124 fish species were grafted onto the chronogram according to their taxonomic affinity to obtain a phylogenetic tree including 498 species. Finally we also present the associated SST data for the observed period (1961–1980) and for the middle (2040–2059) and the end of the 21st century (2080–2099) obtained from NEMOMED8 according to the IPCC A2 scenario. The FishMed database might be of interest in the context of global anthropogenic changes as coastal Mediterranean ecosystems are currently recognized as one of the most impacted ecosystems on earth.

Knowing the past to forecast the future: a case study on a relictual, endemic species of the SW Alps, Berardia subacaulis

Future climate change may lead to a substantial loss of biodiversity, particularly affecting mountain regions, including the Alps. Range-size reduction in high mountain plant species is predicted to be more pronounced for endemic species. Investigating the broad temporal spectrum of range shifts is important for the conservation of biodiversity, since learning how species responded to climate change in the past provides useful insights on how they might react to warming trends in the present and future. Using species distribution models and an ensemble forecasting approach, we explored how the distribution of Berardia subacaulis, a monospecific genus endemic of the south-west Alps, may be affected by past and future projected climate change. During the last interglacial, the habitat suitability of Berardia was lower than present and a progressive increase was observed from the last glacial maximum until now. In the future, Berardia appears to lose more than 80 % of its range, becoming endangered by 2050. Our results suggest that Berardia probably survived past warmer periods in situ, expanding its distributional range during cooler periods. The severe future range contraction predicted for Berardia reflects similar results for other endemic species. As Berardia represents an interesting model species to evaluate the effects of climate warming on range size and shifts, demographic and precise range monitoring may be undertaken on this species.

Intermediate‐term forecasting of aftershocks from an early aftershock sequence: Bayesian and ensemble forecasting approaches

AbstractBecause aftershock occurrences can cause significant seismic risks for a considerable time after the main shock, prospective forecasting of the intermediate‐term aftershock activity as soon as possible is important. The epidemic‐type aftershock sequence (ETAS) model with the maximum likelihood estimate effectively reproduces general aftershock activity including secondary or higher‐order aftershocks and can be employed for the forecasting. However, because we cannot always expect the accurate parameter estimation from incomplete early aftershock data where many events are missing, such forecasting using only a single estimated parameter set (plug‐in forecasting) can frequently perform poorly. Therefore, we here propose Bayesian forecasting that combines the forecasts by the ETAS model with various probable parameter sets given the data. By conducting forecasting tests of 1 month period aftershocks based on the first 1 day data after the main shock as an example of the early intermediate‐term forecasting, we show that the Bayesian forecasting performs better than the plug‐in forecasting on average in terms of the log‐likelihood score. Furthermore, to improve forecasting of large aftershocks, we apply a nonparametric (NP) model using magnitude data during the learning period and compare its forecasting performance with that of the Gutenberg‐Richter (G‐R) formula. We show that the NP forecast performs better than the G‐R formula in some cases but worse in other cases. Therefore, robust forecasting can be obtained by employing an ensemble forecast that combines the two complementary forecasts. Our proposed method is useful for a stable unbiased intermediate‐term assessment of aftershock probabilities.

Application of multiscale analysis-based intelligent ensemble modeling on airport traffic forecast

The air transport industry crucially depends on traffic forecasting for supporting management decisions. In this study, a singular spectrum analysis (SSA)-based ensemble forecasting modeling approach is proposed. The original air passenger time series is first decomposed into three components: trend, seasonal oscillations, and irregular component. The trend is predicted by generalized regression neural network (GRNN), whereas seasonal oscillations are predicted by radial basis function networks (RNFNs). The empirical results of Hong Kong (HK) air passenger data show a significant improvement of the proposed ensemble method in comparison to other results of competing models.