Probabilistic Ensemble Research Articles

Polarimetric synthetic aperture radar image classification for accurate landcover mapping with advanced encoder–decoder networks using multi-dimensional probabilistic voting ensemble

Polarimetric synthetic aperture radar image classification for accurate landcover mapping with advanced encoder–decoder networks using multi-dimensional probabilistic voting ensemble

Multi-Agent Probabilistic Ensembles With Trajectory Sampling for Connected Autonomous Vehicles

Multi-Agent Probabilistic Ensembles With Trajectory Sampling for Connected Autonomous Vehicles

Quantum-Like Approaches Unveil the Intrinsic Limits of Predictability in Compartmental Models.

Obtaining accurate forecasts for the evolution of epidemic outbreaks from deterministic compartmental models represents a major theoretical challenge. Recently, it has been shown that these models typically exhibit trajectory degeneracy, as different sets of epidemiological parameters yield comparable predictions at early stages of the outbreak but disparate future epidemic scenarios. In this study, we use the Doi-Peliti approach and extend the classical deterministic compartmental models to a quantum-like formalism to explore whether the uncertainty of epidemic forecasts is also shaped by the stochastic nature of epidemic processes. This approach allows us to obtain a probabilistic ensemble of trajectories, revealing that epidemic uncertainty is not uniform across time, being maximal around the epidemic peak and vanishing at both early and very late stages of the outbreak. Therefore, our results show that, independently of the models' complexity, the stochasticity of contagion and recovery processes poses a natural constraint for the uncertainty of epidemic forecasts.

Practical Probabilistic Model-Based Reinforcement Learning by Integrating Dropout Uncertainty and Trajectory Sampling.

This article addresses the prediction stability, prediction accuracy, and control capability of the current probabilistic model-based reinforcement learning (MBRL) built on neural networks. A novel approach to dropout-based probabilistic ensembles with trajectory sampling (DPETS) is proposed, where the system uncertainty is stably predicted by combining the Monte Carlo dropout (MC Dropout) and trajectory sampling in one framework. Its loss function is designed to correct the fitting error of neural networks for more accurate prediction of probabilistic models. The state propagation in its policy is extended to filter the aleatoric uncertainty for superior control capability. Evaluated by several Mujoco benchmark control tasks under additional disturbances and one practical robot arm manipulation task, DPETS outperforms related MBRL approaches in both average return and convergence velocity while achieving superior performance than well-known model-free baselines with significant sample efficiency. The open-source code of DPETS is available at https://github.com/mrjun123/DPETS.

A Probabilistic Descent Ensemble for Malware Prediction Using Deep Learning

INTRODUCTION: Introducing a Probabilistic Descent Ensemble (PDE) approach for enhancing malware prediction through deep learning leverages the power of multiple neural network models with distinct architectures and training strategies to achieve superior accuracy while minimizing false positives. OBJECTIVES: Combining Stochastic Gradient Descent (SGD) with early stopping is a potent approach to optimising deep learning model training. Early stopping, a vital component, monitors a validation metric and halts training if it stops improving or degrades, guarding against overfitting. METHODS: This synergy between SGD and early stopping creates a dynamic framework for achieving optimal model performance adaptable to diverse tasks and datasets, with potential benefits including reduced training time and enhanced generalization capabilities. RESULTS: The proposed work involves training a Gaussian NB classifier with SGD as the optimization algorithm. Gaussian NB is a probabilistic classifier that assumes the features follow a Gaussian (normal) distribution. SGD is an optimization algorithm that iteratively updates model parameters to minimize a loss function. CONCLUSION: The proposed work gives an accuracy of 99% in malware prediction and is free from overfitting and local minima.

Time-Dependent Aftershock Probabilistic Forecasting and Ensemble Model Implementation: Insights from the 2021 Ms 7.4 Maduo Earthquake

Abstract On 22 May 2021, an Ms 7.4 earthquake with a focal depth of 17 km struck the Maduo region of Qinghai province, breaking a 3.8-year quiescence of strong earthquakes (magnitude >7.0) in mainland China. This event has increased stress on the Maqin–Maqu segment of the Kunlunshan fault, heightening the potential for a large earthquake in the region. In this study, we employed the epidemic-type aftershock sequence (ETAS) model and Reasenberg–Jones (R–J) model to fit the aftershock sequence following the mainshock, analyzing the temporal response and its ability to influence future generations. Concurrently, ensemble models were explored to leverage the strengths of both the ETAS and R–J models. Short-term forecasts of the probability and occurrence rate of aftershock events with varying magnitudes were conducted for the next three days. Then statistical methods, that is, receiver-operating characteristic diagrams and information gain, were used to evaluate the unconditional and relative performance. Our findings include that the ETAS model indicates a high decay rate with many aftershocks triggered by previous events, whereas the R–J model shows a normal decay rate with a higher proportion of strong aftershocks. The ETAS and R–J models perform better than random guesses for different aftershock magnitudes, but the ETAS model is somewhat affected by the problem of missing small earthquakes for a short period of time after the mainshock. The ensemble model that chooses the minimum strategy shows promise, especially in the early period, suggesting a reasonable and cautious decision approach should be chosen during the unstable stage. This study highlights the importance of short-term aftershock probability estimation for seismic research and decision-making. In the absence of more accurate models, current analytical approaches within the operational earthquake forecasting framework remain valuable. Continuous testing, feedback, and refinement of forecasting models, along with the development of ensemble models, are essential for enhancing seismic risk assessment and mitigation strategies.

Probabilistic Ensemble Framework for Injury Narrative Classification

Probabilistic Ensemble Framework for Injury Narrative Classification

Insights to key operational questions in forecast-informed dam release operation: case of Hume Dam

ABSTRACT Dam operators decide on dam releases given downstream demands and other factors. Downstream tributaries complicate the problem due to significant uncertainties in their flows. Demands can also be highly uncertain. Incorporating probabilistic ensemble forecasts into the decision-making process can mitigate these uncertainties. We investigate how ensemble flow and demand forecasts can inform Hume Dam releases, in the Murray-Darling Basin, for delivery to Lake Mulwala, about four days downstream. We develop and simulate three forecast-based decision rules governing the releases. The first and second rules are based on the expectations and medians respectively of ensemble forecasts of water excess and shortage, forced by ensemble tributary inflow and demand forecasts. The third rule is based on the risks of an excess or shortage per the ensemble forecasts of water excess and shortage. We analyse the results to answer the following key operational questions: (1) Which decision rule is best for managing Hume Dam releases for water supply and under what conditions? (2) How far ahead of flow and demand forecasts should Hume Dam operators adopt? (3) How does limiting the day-to-day change in Hume Dam releases influence the value of forecasts? (4) What is the value of improved demand forecasts?

Developing probabilistic ensemble machine learning models for home-based sleep apnea screening using overnight SpO2 data at varying data granularity.

This study aims to develop sleep apnea screening models with overnight SpO2 data, and to investigate the impact of the SpO2 data granularity on model performance. A total of 7,718 SpO2 recordings from the SHHS and MESA datasets were used. Probabilistic ensemble machine learning was employed to predict sleep apnea status at three AHI cutoff points: ≥ 5, ≥ 15, and ≥ 30 events/hour. To investigate the impact of data granularity, SpO2 data were aggregated at 30, 60, and 300s. Our models demonstrated good to excellent performance on internal test, with average area under the curve (AUC) values of 0.91, 0.93, and 0.96 for cutoffs ≥ 5, ≥ 15, and ≥ 30 at data granularity of 1s, respectively. Both sensitivity (0.76, 0.84, 0.89) and specificity (0.87, 0.86, 0.90) ranged from good to excellent across three cutoffs. Positive predictive values (PPV) ranged from excellent to fair (0.97, 0.83, 0.66), and negative predictive values (NPV) ranged from low to excellent (0.43, 0.87, 0.98). Model performance on external test slightly dropped compared to internal test, but still achieved good to excellent AUC above 0.80 across all data granularity and all the three cutoffs. Data granularity of 300s led to a reduction in performance metrics across all cutoffs. Our models demonstrated superior performance across all three AHI cutoff thresholds compared to existing large sleep apnea screening models, even when considering varying SpO2 data granularity. However, lower data granularity was associated with decreased screening performance, indicating a need for further research in this area.

Santalum album L. alleviates cardiac function injury in heart failure by synergistically inhibiting inflammation, oxidative stress and apoptosis through multiple components

BackgroundHeart failure (HF) is a complex cardiovascular syndrome with high mortality. Santalum album L. (SAL) is a traditional Chinese medicine broadly applied for various diseases treatment including HF. However, the potential active compounds and molecular mechanisms of SAL in HF treatment are not well understood. MethodsThe active compounds and possible mechanisms of action of SAL were analyzed and validated by a systems pharmacology framework and an ISO-induced mouse HF model.ResultsWe initially confirmed that SAL alleviates heart damage in ISO-induced HF model. A total of 17 potentially active components in SAL were identified, with Luteolin (Lut) and Syringaldehyde (SYD) in SAL been identified as the most effective combination through probabilistic ensemble aggregation (PEA) analysis. These compounds, individually and in their combination (COMB), showed significant therapeutic effects on HF by targeting multiple pathways involved in anti-oxidation, anti-inflammation, and anti-apoptosis. The active ingredients in SAL effectively suppressed inflammatory mediators and pro-apoptotic proteins while enhancing the expression of anti-apoptotic factors and antioxidant markers. Furthermore, the synergistic effects of SAL on YAP and PI3K-AKT signaling pathways were further elucidated.ConclusionsMechanistically, the anti-HF effect of SAL is responsible for the synergistic effect of anti-inflammation, antioxidation and anti-apoptosis, delineating a multi-targeted therapeutic strategy for HF.Graphical

Schrödinger as a Quantum Programmer: Estimating Entanglement via Steering

Quantifying entanglement is an important task by which the resourcefulness of a quantum state can be measured. Here, we develop a quantum algorithm that tests for and quantifies the separability of a general bipartite state by using the quantum steering effect, the latter initially discovered by Schrödinger. Our separability test consists of a distributed quantum computation involving two parties: a computationally limited client, who prepares a purification of the state of interest, and a computationally unbounded server, who tries to steer the reduced systems to a probabilistic ensemble of pure product states. To design a practical algorithm, we replace the role of the server with a combination of parameterized unitary circuits and classical optimization techniques to perform the necessary computation. The result is a variational quantum steering algorithm (VQSA), a modified separability test that is implementable on quantum computers that are available today. We then simulate our VQSA on noisy quantum simulators and find favorable convergence properties on the examples tested. We also develop semidefinite programs, executable on classical computers, that benchmark the results obtained from our VQSA. Thus, our findings provide a meaningful connection between steering, entanglement, quantum algorithms, and quantum computational complexity theory. They also demonstrate the value of a parameterized mid-circuit measurement in a VQSA.

Convergence of ensemble forecast distributions in weak and strong forcing convective weather regimes

AbstractThe constraint of computational power and the huge number of degrees of freedom of the atmosphere means a sampling uncertainty exists in probabilistic ensemble forecasts. In our previous study, the uncertainty could be quantified, creating a convergence measure which converges proportional to in the limit of large ensemble size . This power law can then be extrapolated to determine how sampling uncertainty would decrease with larger ensemble sizes and hence find the necessary ensemble size. It is unknown, however, how the sampling uncertainty depends on different weather regimes. This study extends the previous idealised ensemble developed, by including weak and strong forcing convective weather regimes, to look at how sampling uncertainty convergence differs in each. Two ‐member ensembles were run, with weak and strong forcing respectively. Comparisons with a kilometre‐scale weather prediction model ensured realistic weak and strong forcing regimes by comparing the rain, convective available potential energy (CAPE), convective adjustment timescale, and distribution shapes throughout the diurnal cycle. Differences in distribution shape between the regimes led to differences in the convergence measure. Large differences in spread between weak and strong forcing runs throughout the hr period led to large differences in sampling uncertainty of the mean and standard deviation, which could be quantified according to well‐known equations. The timing of these differences was case‐dependent. For extreme statistics such as the quantile and for cases where there was precipitation, the moisture variables for the weak forcing case had the largest sampling uncertainty and required the most members for convergence proportional to . This was due to the tails of the weak forcing moisture variables containing the least amount of density. Different ensemble sizes will hence be required depending on whether one is in the weak or strong forcing convective weather regime.

Recurrent neural network plasticity models: Unveiling their common core through multi-task learning

Recurrent neural network models are known to be particularly suitable for data-driven constitutive modeling due to their built-in memory variables. The main challenge preventing their widespread application to engineering materials lies in their excessive need of data for training. Here, we postulate that RNN models of elasto-plastic solids feature a large common core that is shared by all materials of the same class. The common core is complemented by material-specific layers with parameters that vary from material-to-material. After training RNN models for 25 different von Mises materials, we identify the common core of a multi-task model. Furthermore, we demonstrate through ensemble transfer learning that adding a new material to the multi-task model requires datasets that are two to three orders of magnitude smaller than the datasets needed for training an RNN from scratch. In addition, to multi-task learning, we introduce probabilistic ensembles of RNN plasticity models to quantify the epistemic uncertainty. A deep drawing simulation is performed to demonstrate the superior generalization capabilities of RNNs identified via multi-task learning as compared to those obtained through single task training.

A deterministic and probabilistic hybrid model for wind power forecasting based improved feature screening and optimal Gaussian mixed kernel function

Forecasting wind power is crucial to ensuring the electricity system runs steadily. However, wind power forecasting is hampered by the randomness and volatility of wind power data. In this paper, a deterministic and probabilistic decomposition ensemble model based on improved feature screening and optimal Gaussian mixed kernel function is proposed, which mainly includes feature screening, feature learning, intelligent weighted integration and interval prediction modules. First, an improved feature screening (IFS) is proposed to capture the optimal features of the original wind power data, eliminating the interference of redundant features. Then an optimized deep neural network is suggested to learn the outcomes of IFS and further explore the temporal correlation of these features. Then, considering the difference of the contribution of each feature component, the final result is obtained by integrating the preliminary forecast results using intelligent weighted integration (IWI). Finally, an optimal mixed kernel-based Gaussian process regression model is developed for interval prediction, which improves the adaptability of the model. The data sets of two wind farms in Inner Mongolia were used for empirical testing. On the two datasets, the model's root mean square error (RMSE) is 1.1845 and 1.7354, respectively, both of which are lower than the comparison model. The model has the lowest coverage width criterion (CWC) of 0.1499 and 0.1460 on the two datasets, respectively.

Kidney Ensemble-Net: Enhancing Renal Carcinoma Detection Through Probabilistic Feature Selection and Ensemble Learning

Kidney Ensemble-Net: Enhancing Renal Carcinoma Detection Through Probabilistic Feature Selection and Ensemble Learning

Practicability analysis of online deep reinforcement learning towards energy management strategy of 4WD-BEVs driven by dual-motor in-wheel motors

Deep reinforcement learning (DRL) has emerged as a promising approach for optimizing energy management strategies (EMS) in new energy vehicles. Nevertheless, existing studies typically focus on evaluating the performance of a particular algorithm, overlooking critical implementation details and lacking comprehensive analysis of their real-world applicability. In this paper, we mitigate this issue by conducting a thorough practicability analysis of DRL-based EMS methods. First, we theoretically analyze the benefits and limitations of existing DRL-based EMS approaches based on taxonomy towards their practicability. Subsequently, a novel EMS method that leverages model-based DRL algorithms that other researchers typically underestimate is proposed. The method newly introduces an uncertainty-aware model-based algorithm known as Probabilistic Ensembles with Trajectory Sampling (PETS) and is validated utilizing a four-wheel-drive (4WD) battery electric vehicle (BEV). After that, we conduct a comprehensive practicability analysis of three state-of-the-art DRL algorithms considering critical aspects for real-world deployment, e.g., hyperparameter sensitivity and algorithm transferability. The results demonstrate that even though the on-policy DRL achieves better asymptotic rewards and the off-policy DRL possesses better convergence, the proposed model-based DRL, PETS-based EMS, outperforms others regarding superior robustness and promising transferability across different extents of relevance between tasks. Besides, energy consumption results demonstrate that the model-based EMS can achieve a considerable 96.6% optimality compared to the baseline dynamic programming (DP). Thus, motivated by the challenges of applying DRL algorithms to real-world EMS, our systematic investigation and new insights contribute to advancing the practical employment of DRL-based EMS for BEVs.

Quantifying uncertainty in probabilistic volcanic ash hazard forecasts, with an application to weather pattern based wind field sampling

Probabilistic forecasting of volcanic ash dispersion involves simulating an ensemble of realistic event scenarios to estimate the probability of a particular hazard threshold being exceeded. Although the number of samples that make up the ensemble, how they are chosen, and the desired threshold all set the uncertainty of (or confidence in) the estimated exceedance probability, current practice does not quantify and communicate the uncertainty in ensemble predictions. In this study, we use standard statistical methods to estimate the variance in probabilistic ensembles and use this measure of uncertainty to assess different sampling strategies for the wind field, using the example of volcanic ash transport from a representative explosive eruption in Iceland. For stochastic (random) sampling of the wind field, we show how the variance is reduced with increasing ensemble size and how the variance depends on the desired hazard threshold and the proximity of a target site to the volcanic source. We demonstrate how estimated variances can be used to compare different ensemble designs, by comparing stochastic forecasts with forecasts obtained from a stratified sampling approach using a set of 29 Northern European weather regimes, known as Grosswetterlagen (GWL). Sampling wind fields from within the GWL regimes reduces the number of samples needed to achieve the same variance as compared to conventional stochastic sampling. Our results show that uncertainty in volcanic ash dispersion forecasts can be straightforwardly calculated and communicated, and highlight the need for the volcanic ash forecasting community and operational end-users to jointly choose acceptable levels of variance for ash forecasts in the future.

Land cover analysis of PolSAR images using probabilistic voting ensemble and integrated support vector machine

Land cover analysis of PolSAR images using probabilistic voting ensemble and integrated support vector machine

Ensemble Learning-Based Correlation Coefficient Method for Robust Diagnosis of Voltage Sensor and Short-Circuit Faults in Series Battery Packs

Accurate and reliable multi-fault diagnosis of battery packs is crucial to the safe operation of electric vehicles. To this end, this paper proposes a systematically improved Correlation Coefficient (CC) method by utilizing multivariate statistical analysis and Bayesian probability theory under the framework of ensemble learning. Specifically, different window-widths are first selected in an appropriate value range, and the CC signals between cross-cell voltages for each fixed window-width are calculated to create different local sub-models. Then, in each sub-model, an independent component analysis-based fault diagnosis is implemented to obtain the local diagnostic result, and the results of all sub-models are integrated as an Ensemble Fault Probability (EFP) and an Ensemble Contribution Rate (ECR) through Bayesian probabilistic ensemble interface. Once the EFP exceeds its threshold, a fault is considered to be detected and the ECR is subsequently applied along with the cross-cell sensor topology to identify the fault type (short-circuit or sensor fault) and locate the accurate position of the failed battery cell/sensor. In-depth theoretical analysis and sufficient comparative experiments on a real Lithium-ion battery packs test platform demonstrate that the proposed approach becomes more robust and reliable than the conventional CC-based methods, and also has better physical interpretability.

Post-processing quantitative precipitation forecasts using the seasonally coherent calibration model

ABSTRACT Skilful precipitation ensemble forecasts are necessary to produce trustworthy hydrologic predictions. Raw quantitative precipitation forecasts (QPFs) from the numerical weather prediction (NWP) models are known to be error-prone. In this study, sub-basin averaged deterministic QPFs with five-day lead times from the European Centre for Medium-Range Weather Forecasts (ECMWF) are post-processed through the Seasonally Coherent Calibration (SCC) model for the Narmada and Godavari River basins of India. The SCC model incorporates seasonal climatology from long observations into forecasts and produces calibrated forecasts based on a joint probability model. The SCC model results are compared with the post-processed forecasts from the state-of-the-art Quantile Mapping (QM) method. The results suggest that the probabilistic ensemble forecasts generated from the SCC model have improved skill throughout five-day lead times. Further, the application of SCC-calibrated precipitation forecasts is demonstrated using the Soil & Water Assessment Tool (SWAT) to generate streamflow forecasts.