The ensemble approach to forecasting: A review and synthesis

Wu, Hao; Levinson, David

doi:10.1016/j.trc.2021.103357

Cited by 43 publications

(22 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such an analysis will demonstrate if and to what extent performance varies not only across forecast dates, horizons, and locations, but also across seasons. The results of this analysis may help forecasters and hub coordinators understand sensitivity of contributions to seasonal dynamics, and to reinforce the utility of consortium ensemble approaches that have previously been shown to improve forecasting accuracy (Wang et al 2022; Wu and Levinson 2021).…”

Section: Introductionmentioning

confidence: 63%

Evaluating Performance of Near-Term Influenza Forecasters Across Consecutive Seasons

Nagraj,

Williams,

Turner

2023

Preprint

View full text Add to dashboard Cite

Near-term forecasting efforts for seasonal influenza (flu) aim to enable better public health preparedness before, during, and after each season. The FluSight initiative has fostered flu forecasting activities since 2013. In 2021-22, the organizers switched the primary target to incident weekly flu hospitalizations at state, territorial, and national levels in the United States. Here we studied the performance of contributors who submitted forecasts for this target during the 2021-22 and 2022-23 seasons. We found that forecasters generally did not perform consistently across locations within seasons. For the select group of forecasters who submitted to both seasons, the overall performance relative to one another was not always consistent either. However, several forecasters, including the FluSight ensemble, were among the top performers in both seasons.

show abstract

Section: Introductionmentioning

confidence: 63%

Evaluating Performance of Near-Term Influenza Forecasters Across Consecutive Seasons

Nagraj,

Williams,

Turner

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In addition, the ensemble model, by collective decision mechanism, focuses on synthesizing information from several sub-models with different structures and has been shown to reduce average error and combine the strengths of models in the exploration of diverse data patterns [52][53][54]. However, the addition of a poorly performing model will not reduce the overall model classification skill, because the ensemble model has a net gain compared to the single models [55,56]. Given the above, the ensemble model can reduce the risk of relying on a single prediction distribution and extract richer semantic feature information than the single CNN models (such as each sub-model in the training process has a different probability for boundary regions in pixel-level), which are beneficial in classification tasks to or the achievement of better performance to improve classification accuracy [57][58][59][60][61][62].…”

Section: Discussionmentioning

confidence: 99%

Convolutional neural network-based automatic classification for incomplete antibody reaction intensity in solid phase anti-human globulin test image

Duan

Wang

et al. 2022

Med Biol Eng Comput

View full text Add to dashboard Cite

The precise classification of incomplete antibody reaction intensity (IARI) in hydrogel chromatography medium high density medium solid-phase Coombs test is essential for haemolytic disease screening. However, an automatic and contactless method is required for accurate classification of IARI. Here, we present a deep ensemble learning model that integrates five different convolutional neural networks into a single model for IARI classification. A dataset, including 1628 IARI images and corresponding labels of IARI categories ((-), (1 +), (2 +), (3 +), and (4 +)), was used. We trained our model using 1302 IARIs and validated its performance using 326 IARIs. The proposed model achieved 100%, 99.4%, 99.4%, 100%, and 100% accuracies in the ( −), (1 +), (2 +), (3 +), and (4 +) categories, respectively. The results were compared with those of manual classification by immunologists (average accuracy: 99.8% vs. 88.3%, p < 0.01). Following model assistance, all three immunologists achieved increased accuracy (average accuracy: + 6.1%), with the average accuracy of junior immunologists maximum increasing by 11.3%. The time required for model classification was 0.094 s·image –1 , whereas that required manually was 5.528 s·image –1 . The proposed model can thus substantially improve the accuracy and efficiency of IARI classification and facilitate the automation of haemolytic disease screening equipment. Graphical abstract Supplementary Information The online version contains supplementary material available at 10.1007/s11517-022-02523-1.

show abstract

“…Heterogeneous ensemble learning combines multiple different base models to obtain higher predictive performance than any of the constituent base models alone. One popular way to combine base models is a stacking method (or “super learner”) (Wu and Levinson, 2021; Zhang and Ma, 2012). The steps are as follows.…”

Section: Predictive Modelsmentioning

confidence: 99%

“…While bagging trains base models independently from each other in parallel (Bühlmann and Yu, 2002), boosting does them one-by-one sequentially in an adaptive way, where successive models are trained to predict the remaining error from the previous model (Zhang and Ma, 2012). On the other hand, heterogeneous ensemble uses stacking method (also known as "super learner" method), where a meta-learner is trained on how to combine the output of the heterogeneous base models (Wu and Levinson, 2021). The stacking method differs from the weighted average methods, in that weights for different base models do not have to add up to one.…”

Section: Visuals and Speechmentioning

confidence: 99%

Predicting crowdfunding success with visuals and speech in video ads and text ads

Al-Qershi

Kwon

Zhao

et al. 2022

EJM

View full text Add to dashboard Cite

Purpose For the case of many content features, This paper aims to investigate which content features in video and text ads more contribute to accurately predicting the success of crowdfunding by comparing prediction models. Design/methodology/approach With 1,368 features extracted from 15,195 Kickstarter campaigns in the USA, the authors compare base models such as logistic regression (LR) with tree-based homogeneous ensembles such as eXtreme gradient boosting (XGBoost) and heterogeneous ensembles such as XGBoost + LR. Findings XGBoost shows higher prediction accuracy than LR (82% vs 69%), in contrast to the findings of a previous relevant study. Regarding important content features, humans (e.g. founders) are more important than visual objects (e.g. products). In both spoken and written language, words related to experience (e.g. eat) or perception (e.g. hear) are more important than cognitive (e.g. causation) words. In addition, a focus on the future is more important than a present or past time orientation. Speech aids (see and compare) to complement visual content are also effective and positive tone matters in speech. Research limitations/implications This research makes theoretical contributions by finding more important visuals (human) and language features (experience, perception and future time). Also, in a multimodal context, complementary cues (e.g. speech aids) across different modalities help. Furthermore, the noncontent parts of speech such as positive “tone” or pace of speech are important. Practical implications Founders are encouraged to assess and revise the content of their video or text ads as well as their basic campaign features (e.g. goal, duration and reward) before they launch their campaigns. Next, overly complex ensembles may suffer from overfitting problems. In practice, model validation using unseen data is recommended. Originality/value Rather than reducing the number of content feature dimensions (Kaminski and Hopp, 2020), by enabling advanced prediction models to accommodate many contents features, prediction accuracy rises substantially.

show abstract

The ensemble approach to forecasting: A review and synthesis

Cited by 43 publications

References 51 publications

Evaluating Performance of Near-Term Influenza Forecasters Across Consecutive Seasons

Evaluating Performance of Near-Term Influenza Forecasters Across Consecutive Seasons

Convolutional neural network-based automatic classification for incomplete antibody reaction intensity in solid phase anti-human globulin test image

Predicting crowdfunding success with visuals and speech in video ads and text ads

Contact Info

Product

Resources

About