The prediction of low visibility is essential for proactive traffic safety management on freeways under fog conditions. However, few studies have developed prediction models for visibility on freeways at a short‐term time interval. This study proposes an ensemble learning approach to develop a short‐term prediction model of low visibility on freeways using meteorological data. Spearman's rank correlation coefficient is used to select meteorological variables related to low visibility. Random forests (RF) and extreme gradient boosting (XGB) are employed to develop visibility prediction models, and back propagation neural network (BPNN) and logistic regression (LR) are used for comparison. The models are evaluated over five prediction time intervals (5, 10, 15, 30, and 60 min). The results indicate that the RF models outperform the other models with precision of 73.9%, recall of 59.8% and F1 score of 0.65. Moreover, the prediction model with a 15‐min time interval shows better performance. With the proposed short‐term prediction of low visibility, it is expected that more crashes could be prevented with more appropriate proactive traffic safety management strategies.
Objectives The Society of Thoracic Surgeons (STS), and EuroSCORE II (ES II) risk scores, are the most commonly used risk prediction models for adult cardiac surgery post-operative in-hospital mortality. However, they are prone to miscalibration over time, and poor generalisation across datasets and their use remain controversial. It has been suggested that using Machine Learning (ML) techniques, a branch of Artificial intelligence (AI), may improve the accuracy of risk prediction. Despite increased interest, a gap in understanding the effect of dataset drift on the performance of ML over time remains a barrier to its wider use in clinical practice. Dataset drift occurs when a machine learning system underperforms because of a mismatch between the dataset it was developed and the data on which it is deployed. Here we analyse this potential concern in a large United Kingdom (UK) database. Methods: A retrospective analyses of prospectively routinely gathered data on adult patients undergoing cardiac surgery in the UK between 2012-2019. We temporally split the data 70:30 into a training and validation subset. ES II and five ML mortality prediction models were assessed for relationships between and within variable importance drift, performance drift and actual dataset drift using temporal and non-temporal invariant consensus scoring, combining geometric average results of all metrics as the Clinical Effective Metric (CEM). Results: A total of 227,087 adults underwent cardiac surgery during the study period with a mortality rate of 2.76%. There was a strong evidence of decrease in overall performance across all models (p < 0.0001). Xgboost (CEM 0.728 95CI: 0.728-0.729) and Random Forest (CEM 0.727 95CI 0.727-0.728) were the best overall performing models both temporally and non-temporally. ES II perfomed worst across all comparisons. Sharp changes in variable importance and dataset drift between 2017-10 to 2017-12, 2018-06 to 2018-07 and 2018-12 to 2019-02 mirrored effects of performance decrease across models. Conclusions: Combining the metrics covering all four aspects of discrimination, calibration, clinical usefulness and overall accuracy into a single consensus metric improved the efficiency of cognitive decision-making. All models show a decrease in at least 3 of the 5 individual metrics. CEM and variable importance drift detection demonstrate the limitation of logistic regression methods used for cardiac surgery risk prediction and the effects of dataset drift. Future work will be required to determine the interplay between ML and whether ensemble models could take advantage of their respective performance advantages.
Objectives There is an ongoing debate over whether a procedural specific (e.g. Society of Thoracic Surgeons (STS)) or universal model (e.g. EuroSCORE II (ES II)) should be used for patient selection in cardiac surgery. Recently, we showed that ES II suffers from severe performance drift across several important metrics and that ML approaches such as Xgboost and Random Forest are substantially more resistant to dataset drift. With the growing interest in big data and its leverage through the use of ML approaches that are not limited by linear statistical assumptions, the number of clinical variables can theoretically increase exponentially. In addition, the variations and residual confounding that historically hindered the usefulness of cardiac risk stratification scores can potentially be taken into account. Here, we assess these possibilities on a large United Kingdom (UK) database. Methods: A retrospective analysis of prospectively routinely gathered data on adult patients undergoing cardiac surgery in the UK between 2012-2019. We temporally split the data 70:30 into a training and validation subset. Two sets of seven ML mortality prediction models, with and without variable selection were assessed for consensus Clinical Effective Metric (CEM) overall performance and performance within each of CEM's consistuent metrics. Confounding and potential causal relationships between covariates and outcomes were evaluated using bayesian network analysis. Results: A total of 227,087 adults underwent cardiac surgery during the study period with a mortality rate of 2.76%. For non-variable selected (NVS) risk scores with 102 variables, Xgboost with adjustment for hospital variation was superior to the Xgboost without adjustment (p < 2e-16). Both NVS and the 18 variables selected (VS) Xgboost with adjustment for hospital variation risk scores were superior to the Xgboost (ES II 18 variables) model (p < 6.3e-15), with NVS Xgboost with adjustment for hospital variation having the best performance, followed by the VS Xgboost with adjustment for hospital variation (CEM Difference: 0.0150 and 0.0023, respectively). Conclusions: We have identified an ML adjusted risk score comprising 102 variables that increases risk stratification performance on hold out dataset, removing the need to perform variable selection and reduction. This paves the way for further research that utilises this new set of variables with hospital-based adjustments for the safer selection of patients undergoing cardiac surgery.
How to realize the game equilibrium between bus and nontransit vehicle is a hot topic in the field of transit signal priority (TSP). To this end, a collaborative transit signal priority (Co-TSP) method is proposed. The core of Co-TSP is a two-objective optimization problem which takes the expected delays of buses and the average delays of nontransit vehicles as the objectives. Different from previous studies, Co-TSP uses game theory to realize collaborative optimization, instead of transforming the problem into a single objective optimization problem by weighting. A finite state machine-based algorithm is developed to estimate the average delays of nontransit vehicles. The stochasticity of bus arrival time is also considered in the estimation of bus delays to improve the robustness. Candidate timing plans obtained by the nondominated sorting genetic algorithm (NSGA) are divided into three priority levels based on the delays of buses. The final timing plans can be picked intuitively from the candidates by rules representing expert knowledge and demands to control the priority level. Co-TSP guarantees theoretically by preliminary screening that the expected delays of bus after optimization must be no higher than that before optimization. Simulation experiments are conducted in Shanghai, China, to verify the performance. Results show that Co-TSP reduces the delays of buses by 27.7%∼41.0% and still performs well under low and high congestion levels, while the conventional TSP (CTSP) fails in some cases. Priority control proves to be effective at last. The research provides a new idea for the benefit allocation among participants at intersections.
Speed and punctuality are essential to the quality of bus services. To reduce bus delays and increase bus speed, a transit signal priority (TSP) method based on speed guidance and coordination among consecutive intersections is proposed. The TSP problem is formulated as a binary mixed integer nonlinear program. Impacts of TSP on the current intersection and adjacent intersection downstream are measured by deviations of split time from background timing plans and non-overlapping degrees, respectively. The weighted sum of the two measurements and bus travel time is taken as the objective function. The method does not change the original cycle length, and it is adaptive to timing plans with an arbitrary number of phases. Exclusive bus lanes are required to provide good conditions for speed guidance. A simulation case study of three consecutive intersections in Shanghai, China, is conducted. In the experiments, no priority method, the conventional TSP method, and the proposed method are applied. The results indicate that the proposed method performs the best. Compared to no priority method, the average travel time of buses, delays of bus, and delay per capita are reduced by 26.3%, 91.3%, and 14.5%, respectively. In addition, no serious deterioration is observed in the experience of other road users as the congestion level rises. The study illustrates the possibility of giving high priority to buses without significant negative impacts on other road users, and it can help traffic managers to alleviate traffic congestion in densely populated cities.
Objective The introduction of new clinical risk scores (e.g. European System for Cardiac Operative Risk Evaluation (EuroSCORE) II) superseding original scores (e.g. EuroSCORE I) with different variable sets typically result in disparate datasets due to high levels of missingness for new score variables prior to time of adoption. Little is known about the use of ensemble learning to incorporate disparate data from legacy scores. We tested the hypothesised that Homogenenous and Heterogeneous Machine Learning (ML) ensembles will have better performance than ensembles of Dynamic Model Averaging (DMA) for combining knowledge from EuroSCORE I legacy data with EuroSCORE II data to predict cardiac surgery risk. Methods Using the National Adult Cardiac Surgery Audit dataset, we trained 12 different base learner models, based on two different variable sets from either EuroSCORE I (LogES) or EuroScore II (ES II), partitioned by the time of score adoption (1996–2016 or 2012–2016) and evaluated on holdout set (2017–2019). These base learner models were ensembled using nine different combinations of six ML algorithms to produce homogeneous or heterogeneous ensembles. Performance was assessed using a consensus metric. Results Xgboost homogenous ensemble (HE) was the highest performing model (clinical effectiveness metric (CEM) 0.725) with area under the curve (AUC) (0.8327; 95% confidence interval (CI) 0.8323–0.8329) followed by Random Forest HE (CEM 0.723; AUC 0.8325; 95%CI 0.8320–0.8326). Across different heterogenous ensembles, significantly better performance was obtained by combining siloed datasets across time (CEM 0.720) than building ensembles of either 1996–2011 ( t-test adjusted, p = 1.67×10−6) or 2012–2019 ( t-test adjusted, p = 1.35×10−193) datasets alone. Conclusions Both homogenous and heterogenous ML ensembles performed significantly better than DMA ensemble of Bayesian Update models. Time-dependent ensemble combination of variables, having differing qualities according to time of score adoption, enabled previously siloed data to be combined, leading to increased power, clinical interpretability of variables and usage of data.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
334 Leonard St
Brooklyn, NY 11211
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.