Conditional Probability Model Research Articles

Statistical Metamodel of Liner Acoustic Impedance Based on Neural Network and Probabilistic Learning for Small Datasets

The main novelty of this paper consists of presenting a statistical artificial neural network (ANN)-based model for a robust prediction of the frequency-dependent aeroacoustic liner impedance using an aeroacoustic computational model (ACM) dataset of small size. The model, focusing on percentage of open area (POA) and sound pressure level (SPL) at a zero Mach number, takes into account uncertainties using a probabilistic formulation. The main difficulty in training an ANN-based model is the small size of the ACM dataset. The probabilistic learning carried out using the probabilistic learning on manifolds (PLoM) algorithm addresses this difficulty as it allows constructing a very large training dataset from learning the probabilistic model from a small dataset. A prior conditional probability model is presented for the PCA-based statistical reduced representation of the frequency-sampled vector of the log-resistance and reactance. It induces some statistical constraints that are not straightforwardly taken into account when training such an ANN-based model by classical optimizations methods under constraints. A second novelty of this paper consists of presenting an alternate solution that involves using conditional statistics estimated with learned realizations from PLoM. A numerical example is presented.

A behavioral conditional diffusion probabilistic model for human motion modeling in multi-action mixed human-robot collaboration

Collision avoidance is priority for human-robot collaboration (HRC). Human motions have stochastic, making it difficult for robots to recognize humans. To establish accurate human motion models is important to recognize humans. In the collaboration process, the action categories may suddenly and continuously switch to adapt to complex production tasks, namely, the stochastic of human motions, and we call this collaboration process as multi-action mixed HRC. The previous HRC rarely considered the stochastic of human motions. However, the stochastic in multi-action mixed HRC brings great challenges to human motion modeling. On the one hand, the same observed motions may develop into different action categories, the motion distribution need to be predicted. On the other hand, the connected motions between two action categories are missing, and need to be predicted. To address these two problems, this study introduced diffusion probabilistic models to capture the stochastic of human motions. However, the current diffusion probabilistic models cannot realize the high-accuracy human motion prediction. Most of them did not consider the condition distribution in diffusion process, and did not comprehensively capture the spatiotemporal characteristics between human and robot. To solve the above problems, a behavioral conditional probabilistic diffusion model (BCDPM) is proposed. The BCDPM model can capture the stochastic of human motions to model the human motion distribution. A mask mechanism is combined with diffusion prior generated by BCDPM model to predict the missing connected motions. The BCDPM model consists of the designed constrain network, the denoising network, and enhanced memory graph neural network. The constrain network can generate the condition distribution, the denoising network and enhanced memory graph neural network can capture the short term and long term spatiotemporal characteristics, respectively. Moreover, the overlapping motion primitives are explicitly encoded into the implicit BCDPM model to mine commonalities of motions. The results show the proposed method can obtain excellent accuracy in prediction and outstanding F1 score in collision detection experiment.

Performance-based design of environmental parameters for offshore wind turbine foundations

Taking into account the coupling effects of wind, wave, and current on offshore wind turbines (OWTs), a combined approach utilizing non-linear finite element analysis (FEA), back propagation neural network (BPNN), and probabilistic statistical methodology (PSM) is proposed for accurate environmental parameter estimation in OWT foundation design. The structural bearing performances of OWT foundations are precisely predicted through a combination of FEA and BPNN models, and then serve as constraints for probabilistic statistical analysis. Furthermore, for PSM, a robust multivariate joint probability distribution is constructed leveraging copula functions, which not only maximizes marginal information but also effectively captures the intricate dependencies among ocean environmental parameters. Subsequently, the proposed models, along with traditional univariate analysis, are utilized to estimate the return values of diverse load combinations. The results indicate that, for a given return period, the conditional probability model yields lower wave heights and current velocities compared to univariate analysis, while the joint probability model results in lower values for all ocean parameters, making both approaches more suitable for OWT foundation design. This coupled approach presents a more rational and cost-effective solution compared to univariate analysis, potentially resulting in decreased investment costs.

Study on the distribution of average wind speeds at a mountainous bridge site for structural durability design

Conventional wind speed distribution methods (e.g., Rayleigh distribution and Weibull distribution) may not adequately capture complex characteristics of wind fields in mountainous areas. To address this problem, this study proposes a semi-parametric mix method for modeling the distribution of average wind speeds based on the combination of nonparametric Kernel Density Estimation (KDE) and Generalized Pareto Distribution (GPD). In the proposed method, KDE focuses on capturing the distribution in the main part of average wind speeds, while GPD aims at performing the distribution in terms of those in the extreme part. The segment point (i.e., the threshold) between KDE and GPD distributions is determined based on the combination of conditional mean excesses criterion and empirical rule. Meanwhile, the selection of modeling parameters should ensure that the mix distribution model is continuous and differentiable at the identified threshold point. Then, the commonly-used conditional probability model is further introduced to describe the wind direction distribution. Finally, a case study based on the measured 10-min average wind speeds at a mountainous bridge site is employed to demonstrate the effectiveness of the proposed method. The results indicate that: (1) the distribution of omnidirectional average wind speeds in the mountainous bridge site exhibits an obviously single-peak characteristic, while those considering wind directionality present a certain bimodal characteristic; (2) the proposed method can effectively describe wind speed distributions with different statistical characteristics, and the fitting accuracy outperforms the frequently-employed Weibull distribution model.

Research On the Role and Influencing Factors of Momentum in Tennis Matches Based on Linear Conditional Probability Model and Support Vector Machine

This research investigates the role and influencing factors of momentum in tennis matches using a Linear Conditional Probability Model (LCPM) and Support Vector Machine (SVM). The study begins with data preprocessing to address outliers and standardize scoring notations, followed by an analysis of the momentum's role in tennis, which is defined in relation to consecutive points, games won, and service breaks. The principal component analysis (PCA) is employed to synthesize key factors reflecting momentum, including consecutive scores, untouchable shots, aces, and net points, with a focus on their Markovian properties. A momentum formula is constructed, accounting for the sliding average of scoring differentials and the mutual cancellation of momentum between players. The study further examines the stochasticity of swings in play and runs of success, challenging the notion that these elements are random. The results of the Wald Wolfowitz Run Test and Granger Causality Test indicate that momentum significantly influences the probability of winning points and is not a random phenomenon. A PCA-SVM-SHAP model is developed to predict momentum fluctuations, achieving an 81.26% accuracy rate in the validation set. The model identifies net skills, serving skills, and untouchable shots as the most influential factors on momentum changes. The research extends the model's application to unknown tennis matches and other sports events, demonstrating its generalization ability with varying degrees of accuracy.

Evaluating vegetation vulnerability under compound dry and hot conditions using vine copula across global lands

With global warming, climate extremes are intensifying, particularly compound dry and hot events (CDHEs), which would severely impact vegetation growth. However, quantitatively evaluating vegetation vulnerability to CDHEs remains challenging in global lands. A conditional probability model was constructed using the vine copula function to evaluate responses of vegetation to CDHEs based on the Standardized Moisture Anomaly Index (SZI), the Standardized Temperature Index (STI), and the kernel Normalized Difference Vegetation Index (kNDVI). The results show that vegetation growth is more susceptible to dry and hot conditions in arid and semi-arid regions than in sub-humid and humid regions. In terms of vegetation types, shrubs, savanna, grass, and crops are more susceptible to CDHEs than to individual dry or hot events, especially in water-limited regions. In arid regions, the vegetation vulnerability for open shrubs, closed shrubs, savanna, and grass under extreme CDHEs would correspondingly increase by 45%, 23%, 15%, and 12% compared with dry conditions, indicating the necessity to consider both dry and hot conditions simultaneously when evaluating vegetation dynamics to climatic extremes. In semi-arid and sub-humid regions, broadleaf forests show higher vulnerability to CDHEs than needleleaf forests. However, evergreen broadleaf forests show the lowest vulnerability and higher resistance to CDHEs in humid regions. The Analysis of Variance (ANOVA)-based uncertainty decomposition results show that copula combination sets in vine copula trees and vine copula structures are the two main uncertainty sources, accounting for about 60%. Despite with uncertainty, the modeling results show consistent vegetation loss probability patterns simulated with different structures of the vine copula model, indicating the robustness and reliability of the results. The findings provide a profound comprehension of ecosystems’ response mechanisms to compound extremes, which are essential to developing effective mitigation and adaptation strategies against the impact of future climate change.

A Bayesian Deep Learning-Based Probabilistic Risk Assessment and Early-Warning Model for Power Systems Considering Meteorological Conditions

The ongoing process of decarbonizing the power system and the frequent occurrence of extreme weather have led to a significant increase in operating uncertainty and greatly reduced the system controllability. An effective risk assessment and early-warning tool will significantly assist system operators in monitoring and controlling power systems. However, existing methods mainly rely on simplified analytical conditional probability models to describe the component failure probability under different operating conditions and are not suitable for low-probability risk assessment. Inspired by the idea of Bayesian statistics, a Bayesian deep learning-based probabilistic risk assessment and early-warning (BDL-PRAEW) model considering meteorological conditions is proposed in this paper. A new Bayesian neural network (BNN) is proposed which efficiently utilizes expert experience and domain knowledge as prior to model the contingency probability. And a hybrid neural network is developed to rationally utilize the multi-source heterogeneous data and comprehensively analyze the historical and forecast information. Finally, a novel risk assessment and early-warning model for high-impact, low-probability (HILP) extreme events is proposed, which can predict the risk for the next n time-steps. The proposed method is tested on a modified IEEE 118-bus system and a real-world provincial grid, and the results verify its effectiveness.

Simulation of transportation of acute stroke patients in border regions

Determining the optimal transportation for each stroke patient is critically important to achieve the best possible outcomes. In border regions the next comprehensive stroke center may be just across an international border, but bureaucratic and financial hurdles may prevent a simple transfer to the next stroke center. We hypothesized that in regions close to international borders, patients may benefit from an "open border, closed transfer scenario", meaning that patients in whom a large vessel occlusion (LVO) is detected in the primary stroke center will benefit from a transfer to the nearest stroke center offering endovascular thrombectomy—even if this may be across a national border. We used the Swiss-German–French trinational region as an example for a region with several international borders within close proximity to one another, and compared two feasible scenarios; (a) a “closed borders, open transfer” scenario, where the patient is transported to any center in the same country, (b) an “open border, closed transfer” scenario, where patients are always transported to the nearby primary stroke center first and then to the nearest comprehensive stroke center in either the same or a neighboring country and (c) and “open borders, open transfer” scenario. The outcome of interest was the predicted probability of acute ischemic stroke patients to achieve a good outcome using a conditional probability model which predicts the likelihood of excellent outcome (modified Rankin scale score of 0–1 at 90 days post-stroke) for patients with suspected LVO. Results were modeled in a virtual map from which the ideal transport concept emerged. For an exemplary LVO stroke patient in Germany, the probability of a good outcome was higher in an open border, closed transfer scenario than with closed borders, open transfer (33.1 vs. 30.1%). Moreover, time to EVT would decrease from 232 min in the first scenario to 169 min in an open border, closed transfer scenario. The catchment area of the University Hospital Basel was almost double the size in an open border, closed transfer scenario compared to closed borders (1674 km2 vs. 2897 km2) and would receive transfers from 3 primary stroke centers in other countries (2 in Germany and 1 in France). Stroke patients showed a higher likelihood of good outcome in the “open border” scenarios without transfer restrictions to a specific healthcare system. This probably has implications for stroke treatment in all border regions where EVT eligible stroke patients may benefit from transport to the closest EVT capable center whenever possible, regardless of whether this hospital is located in the same or a neighboring country/jurisdiction.

Investigation of Landslide Susceptibility Decision Mechanisms in Different Ensemble-Based Machine Learning Models with Various Types of Factor Data

Machine learning (ML)-based methods of landslide susceptibility assessment primarily focus on two dimensions: accuracy and complexity. The complexity is not only influenced by specific model frameworks but also by the type and complexity of the modeling data. Therefore, considering the impact of factor data types on the model’s decision-making mechanism holds significant importance in assessing regional landslide characteristics and conducting landslide risk warnings given the achievement of good predictive performance for landslide susceptibility using excellent ML methods. The decision-making mechanism of landslide susceptibility models coupled with different types of factor data in machine learning methods was explained in this study by utilizing the Shapley Additive exPlanations (SHAP) method. Furthermore, a comparative analysis was carried out to examine the differential effects of diverse data types for identical factors on model predictions. The study area selected was Cenxi, Guangxi, where a geographic spatial database was constructed by combining 23 landslide conditioning factors with 214 landslide samples from the region. Initially, the factors were standardized using five conditional probability models, frequency ratio (FR), information value (IV), certainty factor (CF), evidential belief function (EBF), and weights of evidence (WOE), based on the spatial arrangement of landslides. This led to the formation of six types of factor databases using the initial data. Subsequently, two ensemble-based ML methods, random forest (RF) and XGBoost, were utilized to build models for predicting landslide susceptibility. Various evaluation metrics were employed to compare the predictive capabilities of different models and determined the optimal model. Simultaneously, the analysis was conducted using the interpretable SHAP method for intrinsic decision-making mechanisms of different ensemble-based ML models, with a specific focus on explaining and comparing the differential impacts of different types of factor data on prediction results. The results of the study illustrated that the XGBoost-CF model constructed with CF values of factors not only exhibited the best predictive accuracy and stability but also yielded more reasonable results for landslide susceptibility zoning, and was thus identified as the optimal model. The global interpretation results revealed that slope was the most crucial factor influencing landslides, and its interaction with other factors in the study area collectively contributed to landslide occurrences. The differences in the internal decision-making mechanisms of models based on different data types for the same factors primarily manifested in the extent of influence on prediction results and the dependency of factors, providing an explanation for the performance of standardized data in ML models and the reasons behind the higher predictive performance of coupled models based on conditional probability models and ML methods. Through comprehensive analysis of the local interpretation results from different models analyzing the same sample with different sample characteristics, the reasons for model prediction errors can be summarized, thereby providing a reference framework for constructing more accurate and rational landslide susceptibility models and facilitating landslide warning and management.

Conditional probabilistic encounter-impact model for fish-turbine interactions

Knowledge gaps exist in the collective effort to quantify risks and impacts of fish-turbine interactions. Empirical data and modeling studies have characterized stages of fish approach and pass through hydrokinetic turbines, but there has not been a comprehensive model that quantifies conditional occurrence probabilities of fish approaching and then interacting with a turbine in sequential steps. When calculating conditional probabilities, the probability of each step occurring is, in part, dependent on the occurrence of the previous step in the sequence. We combined acoustic data measurements and when data limited, published probabilities to estimate conditional probabilities of tidal turbine impacts on marine animals. The encounter-impact probability model includes approach, entrainment, and impact phases. The approach phase includes probabilities of entry into a spatial domain and zone of influence, followed by entrainment to the device. Each model component of an encounter includes probabilities of active or passive avoidance. Impacts are defined as fish collisions with a device, blade strikes by a rotating blade, and/or a collision followed by a blade strike. Probabilities are estimated using acoustic data from surveys in Admiralty Inlet, WA USA for Pacific herring (Clupea pallasii) encountering an axial or cross-flow turbine during day or night. Conditional probabilities are additive or multiplicative depending if events are cumulative or sequential. Impact probabilities for collisions and blade strikes are estimated individually and then combined to estimate the probability of overall impacts. Probabilities of occurrence for each model component range from 0.0000242 to 0.40. Impact probabilities for collision followed by a blade strike was lowest with estimates ranging from 0.0000242 to 0.0678. Impact probabilities for blade strike had the highest probabilities ranging from 0.000261 to 0.40. Overall impact probabilities had wide variability from 0.00110 to 0.689. Maximum probabilities occurred in each model component and overall impact probabilities for a cross-flow turbine at night with no active or passive avoidance. Estimates were lowest when probabilities were conditional on sequential events, and when active and passive avoidance was applied for an axial-flow turbine during the day. As expected, conditional probabilities were typically lower than equivalent model component and overall probabilities published in the literature. The encounter-impact conditional probability model can be applied to any marine animal in any environment for any hydrokinetic device. Estimating impact probabilities for Pacific herring in Admiralty Inlet for two device types illustrates utilization of existing data and simultaneously identifies data gaps needed to fully calculate empirical-based probabilities for any specific site-species scenario. As additional monitoring data are collected, accuracy of impact risks due to encounters of marine animals with hydrokinetic devices will increase and be confidently used in establishing regulatory policy.

Semi-equivariant conditional normalizing flows, with applications to target-aware molecule generation

Learning over the domain of 3D graphs has applications in a number of scientific and engineering disciplines, including molecular chemistry, high energy physics, and computer vision. We consider a specific problem in this domain, namely: given one such 3D graph, dubbed the base graph, our goal is to learn a conditional distribution over another such graph, dubbed the complement graph. Due to the three-dimensional nature of the graphs in question, there are certain natural invariances such a distribution should satisfy: it should be invariant to rigid body transformations that act jointly on the base graph and the complement graph, and it should also be invariant to permutations of the vertices of either graph. We propose a general method for learning the conditional probabilistic model, the central part of which is a continuous normalizing flow. We establish semi-equivariance conditions on the flow which guarantee the aforementioned invariance conditions on the conditional distribution. Additionally, we propose a graph neural network architecture which implements this flow, and which is designed to learn effectively despite the typical differences in size between the base graph and the complement graph. We demonstrate the utility of our technique in the molecular setting by training a conditional generative model which, given a receptor, can generate ligands which may successfully bind to that receptor. The resulting model, which has potential applications in drug design, displays high quality performance in the key ΔBinding metric.

A decision support tool for risk-benefit analysis of Japanese encephalitis vaccine in travellers.

During pre-travel consultations, clinicians and travellers face the challenge of weighing the risks verus benefits of Japanese encephalitis (JE) vaccination due to the high cost of the vaccine, low incidence in travellers (~1 in 1million), but potentially severe consequences (~30% case-fatality rate). Personalised JE risk assessment based on the travellers' demographics and travel itinerary is challenging using standard risk matrices. We developed an interactive digital tool to estimate risks of JE infection and severe health outcomes under different scenarios to facilitate shared decision-making between clinicians and travellers. A Bayesian network (conditional probability) model risk-benefit analysis of JE vaccine in travellers was developed. The model considers travellers' characteristics (age, sex, co-morbidities), itinerary (destination, departure date, duration, setting of planned activities) and vaccination status to estimate the risks of JE infection, the development of symptomatic disease (meningitis, encephalitis), clinical outcomes (hospital admission, chronic neurological complications, death) and adverse events following immunization. In low-risk travellers (e.g. to urban areas for <1month), the risk of developing JE and dying is low (<1per million) irrespective of the destination; thus, the potential impact of JE vaccination in reducing the risk of clinical outcomes is limited. In high-risk travellers (e.g. to rural areas in high JE incidence destinations for >2months), the risk of developing symptomatic disease and mortality is estimated at 9.5and 1.4per million, respectively. JE vaccination in this group would significantly reduce the risk of symptomatic disease and mortality (by ~80%) to 1.9and 0.3per million, respectively. The JE tool may assist decision-making by travellers and clinicians and could increase JE vaccine uptake. The tool will be updated as additional evidence becomes available. Future work needs to evaluate the usability of the tool. The interactive, scenario-based, personalised JE vaccine risk-benefit tool is freely available on www.VaxiCal.com.

Lossless Point Cloud Geometry and Attribute Compression Using a Learned Conditional Probability Model

In recent years, we have witnessed the presence of point cloud data in many aspects of our life, from immersive media, autonomous driving to healthcare, although at the cost of a tremendous amount of data. In this paper, we present an efficient lossless point cloud compression method that uses sparse tensor-based deep neural networks to learn point cloud geometry and color probability distributions. Our method represents a point cloud with both occupancy feature and three attribute features at different bit depths in a unified sparse representation. This allows us to efficiently exploit feature-wise and point-wise dependencies within point clouds using a sparse tensor-based neural network and thus build an accurate auto-regressive context model for an arithmetic coder. To the best of our knowledge, this is the first learning-based lossless point cloud geometry and attribute compression approach. Compared with the-state-of-the-art lossless point cloud compression method from Moving Picture Experts Group (MPEG), our method achieves 22.6% reduction in total bitrate on a diverse set of test point clouds while having 49.0% and 18.3% rate reduction on geometry and color attribute component, respectively.

Probabilistic analysis on the influences of heatwaves during the onset of flash droughts over China

Abstract The increasing concurrences of heatwaves and droughts in the context of global warming have attracted much attention from the scientific community given their devastating social and environmental impacts. In this study, the effects of heatwaves in each adjacent week of flash drought onset on the intensification rate of soil moisture were quantified through a meta-Gaussian-based conditional probability model. Results showed that both heatwaves and flash droughts have become more frequent since the middle of the 1990s. For the seasonal distributions, except for the southwestern region where flash droughts lagged behind heatwaves, there was a good synchronization between the two climate extremes. Strong correlations between heatwaves and flash droughts were found in the northeastern, northern, and southwestern regions. Heatwaves with varied timing of emergence behave differently on the formation of flash droughts, along with significant regional differences. Short-term impending hot conditions were crucial for the breakout of flash droughts, especially for the week when flash droughts were initiated, the emergence of heatwaves was likely to increase the intensification rate of soil moisture by 20% compared to those with no heatwaves in their development stage.

Modeling Sequential Annotations for Sequence Labeling With Crowds.

Crowd sequential annotations can be an efficient and cost-effective way to build large datasets for sequence labeling. Different from tagging independent instances, for crowd sequential annotations, the quality of label sequence relies on the expertise level of annotators in capturing internal dependencies for each token in the sequence. In this article, we propose modeling sequential annotation for sequence labeling with crowds (SA-SLC). First, a conditional probabilistic model is developed to jointly model sequential data and annotators' expertise, in which categorical distribution is introduced to estimate the reliability of each annotator in capturing local and nonlocal label dependencies for sequential annotation. To accelerate the marginalization of the proposed model, a valid label sequence inference (VLSE) method is proposed to derive the valid ground-truth label sequences from crowd sequential annotations. VLSE derives possible ground-truth labels from the tokenwise level and further prunes subpaths in the forward inference for label sequence decoding. VLSE reduces the number of candidate label sequences and improves the quality of possible ground-truth label sequences. The experimental results on several sequence labeling tasks of Natural Language Processing show the effectiveness of the proposed model.

An Empirical Modelling and Simulation Framework for Fire Events Initiated by Vegetation and Electricity Network Interactions

Electrical infrastructure is one of the major causes of bushfire in Australia alongside arson and lightning strikes. The two main causes of electrical-infrastructure-initiated fires are asset failure and powerline vegetation interactions. In this paper, we focus on powerline–vegetation interactions that are caused by vegetation falling onto or blowing onto electrical infrastructure. Currently, there is very limited understanding of both the spatio-temporal variability of these events and their causative factors. Bridging this knowledge gap provides an opportunity for electricity utility companies to optimally allocate vegetation management resources and to understand the risk profile presented by vegetation fall-in initiated fires, thereby improving both operational planning and strategic resource allocation. To bridge this knowledge gap, we developed a statistical rare-event modelling and simulation framework based on Endeavour Energy’s fire start and incident records from the last 10 years. The modelling framework consists of nested, rare-event-corrected, conditional probability models for vegetation events and consequent ignition events that provide an overall model for vegetation-initiated ignitions. Model performance was tested on an out-of-time test set to determine the predictive utility of the models. Predictive performance was reasonable with test set AUC values of 0.79 and 0.66 for the vegetation event and ignition event models, respectively. The modelling indicates that wind speed and vegetation features are strongly associated with vegetation events, and that Forest Fire Danger Index (FFDI) and soil type are strongly associated with ignition events. The framework can be used by energy utilities to optimize resource allocation and prepare future networks for climate change.

基于朴素贝叶斯算法评估页岩油藏产能

Shale oil, as an unconventional oil and gas resource with huge reserves, has become an important replacement resource and great significance to develop. Aiming at the problems of rapid decline oil production in the depletion development of shale oil reservoirs by "depleted horizontal well+volume fracturing", and the lower recovery factor predicted by the existing productivity evaluation methods in shale reservoir, we construct a productivity evaluation method for fractured horizontal wells based on naive Bayes algorithm. Taking the Jimsar shale reservoir in Xinjiang oilfield as the target reservoir, we establish the multi-classification Naive Bayes prior probability model and the conditional probability model containing five attributes of geological parameters and engineering parameters by taking the three-year cumulative production as the classification evaluation index. Then we obtain the posterior-probability model of the four types of production capacity based on Bayesian theory to achieve the capacity assessment of fractured horizontal wells in shale oil reservoirs. The results show that the proposed shale reservoir capacity classification method based on the Naive Bayes model is applicable. The accuracy of productivity prediction is 94% for Class Ⅰ wells, 71% for Class Ⅱ wells, 87% for Class Ⅲ wells, and 92% for Class Ⅳ wells, respectively. The productivity classification trend distribution map of fracturing horizontal wells in Jimsar shale reservoir is drawn, and the high-production potential area is mainly distributed in the southeast of the reservoir. The analysis of the constructed posterior probability model shows that the optimization of fracturing construction parameters can improve the probability of high well production rate. The study provides a guiding basis for the subsequent large-scale fracturing reconstruction of horizontal wells for shale oil development.

Modeling optimal patient transport in a stroke network capable of remote telerobotic endovascular therapy.

Telerobotic endovascular therapy (EVT) has the potential to decrease time to treatment and expand existing networks of care to more rural populations. It is currently unclear how its implementation would impact existing stroke networks. Conditional probability models were generated to predict the probability of excellent outcome for patients with suspected large vessel occlusion (LVO). A baseline stroke network was created for California using existing intravenous thrombolysis (IVT) centers and comprehensive stroke centers (CSCs) capable of IVT and EVT. Optimal transport decisions and catchment areas were generated for the baseline model and three hypothetical scenarios through conversion of IVT centers at various distances from a CSC into centers capable of telerobotic EVT [i.e., hospitals ≥15 and <50 miles from a CSC were converted (Scenario 1), ≥50 and <100 miles (Scenario 2), and ≥100 miles (Scenario 3)]. Procedural times and success rates were varied systematically. Telerobotic EVT centers decreased median travel time for LVO patients in all three scenarios. The estimated number of robotically treated LVOs per year in Scenarios 1, 2, and 3 were 2,172, 740, and 212, respectively. Scenario 1 (15-50 miles) was the most sensitive to robotic time delay and success rate, but all three scenarios were more sensitive to decreases in procedural success rate compared to time delay. Telerobotic EVT has the potential to improve care for stroke patients outside of major urban centers. Compared to procedural time delays in robotic EVT, a decrease in procedural success rate would not be well tolerated.

Quantifying Improved Outcomes, Cost Savings, and Hospital Volume Changes From Optimized Emergency Stroke Transport.

A previously published conditional probability model optimizes prehospital emergency transport protocols for patients with suspected large-vessel occlusion by recommending the transport strategy, drip-and-ship or mothership, that results in a higher probability of an excellent outcome. In this study, we create generalized models to quantify the change in annual hospital patient volume, the expected annual increase in the number of patients with an excellent outcome, and the annual cost savings to a single-payer healthcare system resulting from these optimized transport protocols. We calculated the expected number of patients with suspected large-vessel occlusion transported by ambulance over a 1-year period in a region of interest, using the annual stroke incidence rate and a large-vessel occlusion screening tool. Assuming transport to the closest hospital is the baseline transport policy across the region (drip-and-ship), we determined the change in annual hospital patient volume from implementing optimized transport protocols. We also calculated the resulting annual increase in the number of patients with an excellent outcome (modified Rankin Score of 0-1 at 90 days) and associated cost savings to a single-payer healthcare system. We then performed a case study applying these generalized models to the stroke system serving the Greater Vancouver and Fraser Valley Area, BC, Canada. In the Greater Vancouver and Fraser Valley Area, there was an annual increase of 36 patients with an excellent outcome, translating to an annual cost savings of CA$2 182 824 to the British Columbia healthcare system. We also studied how these results change depending on our assumptions of treatment times at the regional stroke centers. Our framework quantifies the impact of optimized emergency stroke transport protocols on hospital volume, outcomes, and cost savings to a single-payer healthcare system. When applied to a specific region of interest, these models can help inform health policies concerning emergency transport of patients with suspected large-vessel occlusion.

Fair Transmission of Individual Signals and Formation of Mainstream Information: Evidence from Herd Behaviours in Emergencies.

Risk society is full of emergencies, accompanied by uncertainties and losses. Under emergencies, controlling herd behaviour is challenging due to more interactions and changes among individuals. This research establishes Bayes conditional probability models to explain the fair transmission of individual signals and individual decision-making after receiving others' signals. The simulation shows the following conclusions: first, each individual has a fair chance to influence the mainstream information; second, the order in which individuals make decisions during an emergency affects the difficulties and likelihood of making a rational decision; third, the high authority of information can become mainstream and guide individual behaviour; and fourth, two individual characteristics, including risk appetite and personal experience, are important in the fair transmission of individual signals and formation of mainstream information. According to the findings, this research proposes two strategies, including interfering with information and controlling existing key opinion leaders to control the mainstream information within a group in emergencies. These two strategies are proved to be useful in detecting and preventing approaches to alleviate individual herd behaviour, which should be monitored and controlled in machine learning models for individual behaviour simulation and prediction. Compared to previous research that focuses on media and public opinion in emergencies, this research focuses on a specific type of information (i.e., individual decision-making and actions) on the individual level and its effects on herd behaviours within the group. This research complements the explanation of the micro-mechanism of how individuals receive information and make decisions and actions.