Location Probability Research Articles

Integrating symbolic reasoning into neural generative models for design generation

Design generation requires tight integration of neural and symbolic reasoning, as good design must meet explicit user needs and honor implicit rules for aesthetics, utility, and convenience. Current automated design tools driven by neural networks produce appealing designs, but cannot satisfy user specifications and utility requirements. Symbolic reasoning tools, such as constraint programming, cannot perceive low-level visual information in images or capture subtle aspects such as aesthetics. We introduce the Spatial Reasoning Integrated Generator (SPRING) for design generation. SPRING embeds a neural and symbolic integrated spatial reasoning module inside the deep generative network. The spatial reasoning module samples the set of locations of objects to be generated from a backtrack-free distribution. This distribution modifies the implicit preference distribution, which is learned by a recurrent neural network to capture utility and aesthetics. The sampling from the backtrack-free distribution is accomplished by a symbolic reasoning approach, SampleSearch, which zeros out the probability of sampling spatial locations violating explicit user specifications. Embedding symbolic reasoning into neural generation guarantees that the output of SPRING satisfies user requirements. Furthermore, SPRING offers interpretability, allowing users to visualize and diagnose the generation process through the bounding boxes. SPRING is also adept at managing novel user specifications not encountered during its training, thanks to its proficiency in zero-shot constraint transfer. Quantitative evaluations and a human study reveal that SPRING outperforms baseline generative models, excelling in delivering high design quality and better meeting user specifications.

Improving Deep Learning Models for Pediatric Low-Grade Glioma Tumours Molecular Subtype Identification Using MRI-based 3D Probability Distributions of Tumour Location.

Purpose: Pediatric low-grade gliomas (pLGG) are the most common brain tumour in children, and the molecular diagnosis of pLGG enables targeted treatment. We use MRI-based Convolutional Neural Networks (CNNs) for molecular subtype identification of pLGG and augment the models using tumour location probability maps. Materials and Methods: MRI FLAIR sequences of 214 patients (110 male, mean age of 8.54 years, 143 BRAF fused and 71 BRAF V600E mutated pLGG tumours) from January 2000 to December 2018 were included in this retrospective REB-approved study. Tumour segmentations (volumes of interest-VOIs) were provided by a pediatric neuroradiology fellow and verified by a pediatric neuroradiologist. Patients were randomly split into development and test sets with an 80/20 ratio. The 3D binary VOI masks for each class in the development set were combined to derive the probability density functions of tumour location. Three pipelines for molecular diagnosis of pLGG were developed: location-based, CNN-based, and hybrid. The experiment was repeated 100 times each with different model initializations and data splits, and the Areas Under the Receiver Operating Characteristic Curve (AUROC) was calculated, and Student's t-test was conducted. Results: The location-based classifier achieved an AUROC of 77.9, 95% confidence interval (CI) (76.8, 79.0). CNN-based classifiers achieved an AUROC of 86.1, 95% CI (85.0, 87.3), while the tumour-location-guided CNNs outperformed the other classifiers with an average AUROC of 88.64, 95% CI (87.6, 89.7), which was statistically significant (P-value .0018). Conclusion: Incorporating tumour location probability maps into CNN models led to significant improvements for molecular subtype identification of pLGG.

Machine learning approach for detection of land subsidence induced by underground coal fire using multi-sensor satellite data

ABSTRACT High-rank coal reserves in Jharia Coalfield (JCF, India), are invariably associated with underground coal fires and land subsidence. This study explores multi-sensor time series satellite data (Landsat 8 OLI and Sentinel-1) through machine learning (ML) to determine the regional ground deformation accompanying coal fires and their contextual relationship. The results show that the highest degree of subsidence is closely associated with the active mine benches with overburden dumps. The relationship between the coal fire and land subsidence parameters is considered as a binary classification problem, explored by calculating the probability of subsidence with a desirable categorical outcome through different ML models. The accuracy of the models is validated using performance metrics that shows that the Random Forest (RF) metrics predict the probability of deformation locations in response to the volume reduction of the burning coal fire and vertical compression due to Overburden Dump (OBD) near active mine benches. The estimated displacement trends have been used to forecast the Autoregressive Integrated Moving Average (ARIMA) method, estimated using Line-of-Sight (LOS) displacement values vary around the best fit within the 95% confidence limits. The trend shows ∼15–25% increase in subsidence compared to the cumulative subsidence.

Structural damage prediction based on Bayesian updating with Monte Carlo sampling

A series of engineering accidents in recent years have caused people to worry about the safety of engineering structures. As buildings continue to age, engineering accidents are increasing, and it is necessary to monitor their health. Bayesian updating can determine the probability of damage to a structure by collecting a posteriori information, and can more accurately estimate the location and extent of damage, thereby predicting the risk of structural damage and repairing the damage in time, significantly improving engineering efficiency. This paper attempts to update the information of the simulated structure based on the Bayesian updating method to obtain a more accurate posterior distribution of the potentially damaged components to demonstrate the feasibility and superiority of the technique. This paper finds that Bayesian updating can be used in conjunction with modeling software such as SMsolver. By constructing a structural model with SMsolver and processing the data with Bayesian updating, it is possible to calculate the damage problems of the structure accurately. At the same time, this paper also uses Monte Carlo sampling technology in combination with the Bayesian updating method to explore the prediction effect of structural damage. Using Monte Carlo sampling methods, the failure probability and damage location of the structure can be effectively estimated under the condition of known prior information, thereby improving production efficiency and the accuracy of structural inspection.

Observations of Compressional Structures Driven by Interaction Between Foreshock Ions and Discontinuities

AbstractThe ion foreshock is very dynamic, characterized by various transient structures that can perturb the bow shock and influence the magnetosphere‐ionosphere system. One important driver of foreshock transients is solar wind directional discontinuities (DDs) that demagnetize foreshock ions leading to a local current. If this current decreases the field strength at the DD, a hot flow anomaly (HFA) can form. Recent hybrid simulations found that when the current increases the field strength at the DD, a compressional structure forms with enhanced density and field strength opposite to HFAs. Using MMS and THEMIS observations, we confirm this situation. We demonstrate that the current geometry driven by the foreshock ions plays a critical role in the formation. The initial gyrophase of foreshock ions, due to their specular reflection, determines whether they can cross the DD. When many of the foreshock ions cannot cross the DD and the local current they drive increases the field strength at the DD, the enhanced field strength inhibits more foreshock ions from crossing the DD, further enhancing the local current. This feedback loop promotes the growth of the compressional structure. Such foreshock ion‐driven compressional structures can result in dynamic pressure enhancements in the magnetosheath, leading to magnetosheath jets. Our study enables prediction of the location and formation probability of such compressional structures and their potential geoeffectiveness.

Strategic Planning of Aerial Assets for Disaster Response

The rapid deployment of fleets of small, uncrewed aircraft (drones) in the immediate aftermath of a natural disaster to search impacted regions for people in need of rescue is one of the most vital applications of advanced air mobility. Effective drone-based search operations require that the drone fleets operate out of bases that are appropriately located in advance of the disaster. Using a case study based in the Iwate prefecture of Japan, we develop optimization formulations to strategically locate drone bases. It is important to be capable of responding quickly to the locations most likely to require a search, while covering as large an area as possible. We evaluate the disparities in the level of access afforded to different areas. We extend our optimization formulation to account for the probability of the base locations themselves being impacted by the disaster and the possibility of base relocation. Finally, we illustrate how a vehicle routing component can be used to address the tactical portion of drone-based search operations.

Peripheral vision contributes to implicit attentional learning: Findings from the "mouse-eye" paradigm.

The central visual field is essential for activities like reading and face recognition. However, the impact of peripheral vision loss on daily activities is profound. While the importance of central vision is well established, the contribution of peripheral vision to spatial attention is less clear. In this study, we introduced a "mouse-eye" method as an alternative to traditional gaze-contingent eye tracking. We found that even in tasks requiring central vision, peripheral vision contributes to implicit attentional learning. Participants searched for a T among Ls, with the T appearing more often in one visual quadrant. Earlier studies showed that participants' awareness of the T location probability was not essential for their ability to learn. When we limited the visible area around the mouse cursor, only participants aware of the target's location probability showed learning; those unaware did not. Adding placeholders in the periphery did not restore implicit attentional learning. A control experiment showed that when participants were allowed to see all items while searching and moving the mouse to reveal the target's color, both aware and unaware participants acquired location probability learning. Our results underscore the importance of peripheral vision in implicitly guided attention. Without peripheral vision, only explicit, but not implicit, attentional learning prevails.

Metabolic regulation of single synaptic vesicle exo- and endocytosis in hippocampal synapses

Glucose has long been considered a primary energy source for synaptic function. However, it remains unclear to what extent alternative fuels, such as lactate/pyruvate, contribute to powering synaptic transmission. By detecting individual release events in hippocampal synapses, we find that mitochondrial ATP production regulates basal vesicle release probability and release location within the active zone (AZ), evoked by single action potentials. Mitochondrial inhibition shifts vesicle release closer to the AZ center and alters the efficiency of vesicle retrieval by increasing the occurrence of ultrafast endocytosis. Furthermore, we uncover that terminals can use oxidative fuels to maintain the vesicle cycle during trains of activity. Mitochondria are sparsely distributed along hippocampal axons, and we find that terminals containing mitochondria display enhanced vesicle release and reuptake during high-frequency trains. Our findings suggest that mitochondria not only regulate several fundamental features of synaptic transmission but may also contribute to modulation of short-term synaptic plasticity.

Integrating Microseismic Monitoring for Predicting Water Inrush Hazards in Coal Mines

The essence of roof water inrush in coal mines fundamentally stems from the development of water-bearing fracture zones, facilitating the intrusion of overlying aquifers and thereby leading to water hazard incidents. Monitoring rock-fracturing conditions through the analysis of microseismic data can, to a certain extent, facilitate the prediction and early warning of water hazards. The water inflow volume stands as the most characteristic type of data in mine water inrush accidents. Hence, we investigated the feasibility of predicting water inrush events through anomalies in microseismic data from the perspective of water inflow volume variations. The data collected from the microseismic monitoring system at the 208 working face were utilized to compute localization information and source parameters. Based on the hydrogeological conditions of the working face, the energy screening range and its calculation grid characteristics were determined, followed by the generation of kernel density cloud maps at different depths. By observing these microseismic kernel density cloud maps, probabilities of roof water-conducting channel formation and potential locations were inferred. Subsequently, based on the positions of these roof water-conducting channels on the planar domain, the extension depth and expansion direction of the water-conducting channels were determined. Utilizing microseismic monitoring data, a quantitative assessment of water inrush risk was conducted, thereby establishing a linkage between microseismic data and water (inrush) data, which are two indirectly related datasets. The height of microseismic events was directly proportional to the trend of water inflow in the working face. In contrast, the occurrence of water inflow events and microseismic events exhibited a specific lag effect, with microseismic events occurring prior to water inrush events. Abnormalities in microseismic monitoring data partially reflect changes in water-conducting channel patterns. When connected with coal seam damage zones, water inrush hazards may occur. Therefore, abnormalities in microseismic monitoring data can be regarded as one of the precursor signals indicating potential floor water inrushes in coal seams.

Water Pollution Detection Using Video Surveillance and Deep Learning

Aiming at the problems that the traditional water quality prediction model is generally not high in prediction accuracy and robustness, a water pollution prediction using deep learning in water environment monitoring big data is proposed. Objective. To optimize and improve the prediction accuracy of the water quality prediction model. Firstly, in the water environment monitoring system, the Internet of Things big data technology is used to accurately sense and monitor the real-time data of sewage treatment equipment and sewage quality. Then, the deep belief network (DBN) is used to build the water pollution prediction model, and the collected sewage treatment data is analyzed to predict the water quality status. Finally, particle swarm optimization algorithm is used to dynamically optimize the number of hidden layer neural units and learning rate in the DBN prediction model, which makes the prediction results more scientific and accurate. Based on the sampling data of Shanghai Jinze Reservoir, the proposed model is experimentally analyzed. The results show that the probability of accurate location of the pollution source is not less than70%. And under the two indicators of chemical oxygen demand and biological oxygen demand, the root mean square error and correlation coefficient are 3.073, 0.9892 and 1.958, 0.9565, respectively, which are better than other comparison models.

Attenuation of spatial bias with target template variation

This study investigated the impact of target template variation or consistency on attentional bias in location probability learning. Participants conducted a visual search task to find a heterogeneous shape among a homogeneous set of distractors. The target and distractor shapes were either fixed throughout the experiment (target-consistent group) or unpredictably varied on each trial (target-variant group). The target was often presented in one possible search region, unbeknownst to the participants. When the target template was consistent throughout the biased visual search, spatial attention was persistently biased toward the frequent target location. However, when the target template was inconsistent and varied during the biased search, the spatial bias was attenuated so that attention was less prioritized to a frequent target location. The results suggest that the alternative use of target templates may interfere with the emergence of a persistent spatial bias. The regularity-based spatial bias depends on the number of attentional shifts to the frequent target location, but also on search-relevant contexts.

Facilitation and interference are asymmetric in holistic face processing

A hallmark of face specificity is holistic processing. It is typically measured by paradigms such as the part–whole and composite tasks. However, these tasks show little evidence for common variance, so a comprehensive account of holistic processing remains elusive. One aspect that varies between tasks is whether they measure facilitation or interference from holistic processing. In this study, we examined facilitation and interference in a single paradigm to determine the way in which they manifest during a face perception task. Using congruent and incongruent trials in the complete composite face task, we found that these two aspects are asymmetrically influenced by the location and cueing probabilities of the target facial half, suggesting that they may operate somewhat independently. We argue that distinguishing facilitation and interference has the potential to disentangle mixed findings from different popular paradigms measuring holistic processing in one unified framework.

Parameter estimation and sensitivities of a Bayesian source localization method in an urban environment

Acoustic signals propagating in urban environments are influenced by rough-surface scattering, multipath reflections, and diffraction. The performance of conventional source localization algorithms often suffers when these effects are present. Bayesian approaches, however, are particularly well suited to incorporating physics-based statistical models for the signal propagation. Previously, we found that the complex Wishart distribution, which describes fully saturated scattered signals across a network of receivers, can be readily employed in a Bayesian framework. In this presentation, a new source localization algorithm based on the Wishart distribution signal model is described and tested on real acoustic data. The experimental data were collected within a mock urban environment as part of a NATO urban acoustics-seismics experiment in Walenstadt, Switzerland. Four acoustic arrays recorded signatures from gunshots and ground vehicles. The present study uses the measured signals to investigate the sensitivity and relative importance of the model parameters, including source and noise amplitudes, frequency band, prior signal sampling, and sensor-sensor correlation behavior. Results are presented in the form of source location probability maps and localization error metrics. Interpretations of the study, including strengths and weaknesses of the new method, are discussed.

Evaluation of TV White Space Implementation in Indonesia: A Case Study in Yogyakarta

TV white space is a technology that utilizes unused spectrum owned by primary users, namely TV broadcasters in a particular area and time. There are two main problems related to the use of white space, namely protection of primary users and identification of white space channels. The prohibited, strict, moderate, and loose (PSML) method has previously been proposed, dividing the TV protection area into four zones to model the white space system. Researchers generally use additional protection areas and television coverage areas with a location probability of less than 70% to consider the possibility of overlap between white space and protected areas. However, this method has not been in implementation in the field. In this study, the author implemented the PSML method in the city of Yogyakarta by using the coverage area of TVRI Patuk as a sample. Based on the measurement results, the author discovered that the deviation between the simulation results and the measurement results was 7.32 dB or 1.82 dB higher than ITU’s recommendation. This deviation indicates that, based on geographical conditions, Yogyakarta has a higher variation in field strength values than ITU’s recommendations.

Numerical methods for fractional Fokker–Planck equation with multiplicative Marcus Lévy noises

The Fokker–Planck equation (FPE) is an important deterministic tool for investigating stochastic dynamical systems. In this paper, we consider the space-time fractional FPE driven by multiplicative Marcus Lévy noises. Efficient numerical schemes are presented to solve the equations. Stability and convergence of the methods are also discussed. We give some numerical experiments to validate our schemes, and examine the effects of parameters on solutions. Additionally, we analyze the maximal likely trajectories and the critical time for the change of the most probability location.

Source term estimation for continuous plume dispersion in Fusion Field Trial-07: Bayesian inference probability adjoint inverse method

In scenarios involving sudden releases of unidentified gases or concealed pollution emergencies, source control emerges as a critical procedure to safeguard residential air quality. Appropriate inverse source tracking methodology depending on diverse measurement data could be utilized to promptly identify pollutant source parameters. In this study, source term estimation (STE) method, i.e., jointly combining probability adjoint method with the Bayesian inference method, has been proposed. General form of the pollutant inverse transport equation was firstly established. Subsequently, the pollution source information, assumed from single continuous point releases during Fusion Field Trials 2007 under an unsteady wind field, was identified using the Bayesian inference probability adjoint inverse method. Metropolis-Hastings Markov Chain Monte Carlo (MH-MCMC) and Differential Evolution Markov Chain Monte Carlo (DE-MCMC) were then compared as sampling methods for Bayesian inference. Results indicated that the DE-MCMC algorithm has superior convergence and could present higher accuracy of pollutant source information than that of MH-MCMC algorithm, particularly for highly nonlinear and multi-modal distribution systems. Furthermore, the integration of Union standard Adjoint Location Probability (UALP) as prior information into the Bayesian inference probability adjoint inverse method effectively narrowed the sampling range, enhancing both the accuracy and robustness of the proposed approach. Finally, the impact of the covariance matrix on the inverse identification accuracy was explored. Overall, this research has provided insights into the future applicability of this Bayesian inference inversion technique for point source identification.

Continual Learning for SAR Target Incremental Detection via Predicted Location Probability Representation and Proposal Selection

Continual Learning for SAR Target Incremental Detection via Predicted Location Probability Representation and Proposal Selection

Producing affordable housing in higher-opportunity neighborhoods: Incentives in California’s LIHTC program

ABSTRACT Housing policy shapes where low-income families live and whether they can access neighborhoods that promote children’s well-being. Since 2018, California’s Low-Income Housing Tax Credit (LIHTC) program has incentivized the development of family affordable housing in higher-resource areas. We examine how the location of proposed and funded LIHTC developments changed in response to these incentives. We find that the probability that large family projects—those eligible for incentives—were proposed in higher-resource areas increased from 0.19 to 0.29, comparing 2014-2017 and 2018–2021. The probability that funded large family projects were located in higher-resource areas doubled, from 0.15 to 0.30. The probability of location in higher-resource areas declined for developments ineligible for incentives. Interviews with affordable housing developers illuminate the role of mission and resources in shaping their responses to program incentives. We conclude with a broader discussion of the LIHTC program as a tool for reducing neighborhood inequality.

A new adaptive multi-kernel relevance vector regression for structural reliability analysis

Surrogate models have been widely used in structural reliability analysis to improve the computational efficiency and the accuracy of failure probability. Recently, several multi-kernel relevance vector regression (MKRVR) models have been studied to evaluate the failure probability. However, existing multiple kernel functions for relevance vector regression models are fixed choices, which increases the number of calls to the limit state function (LSF) and leads to inaccurate results. To address the problem, this paper presents a new adaptive MKRVR model combined with Monte Carlo simulation (MCS). Firstly, a stepwise kernel selection strategy is developed to adaptively select better-performing kernel functions and eliminate redundant kernel functions for constructing the MKRVR model. Secondly, a new active learning function is proposed by considering the probability of mis-prediction and spatial locations of the existing sampling point to identify the new training sample points. Thirdly, a hybrid efficient stopping criterion is adopted to terminate the learning process automatically. Three benchmark examples and one practical engineering example are introduced to demonstrate the effectiveness of the proposed method. Results show that the proposed method can provide accurate failure probability by less number of calls to the LSF than existing fixed kernel-based methods.

Source Location Identification in an Ideal Urban Street Canyon with Time-Varying Wind Conditions under a Coupled Indoor and Outdoor Environment

Source location identification methods are typically applied to steady-state conditions under pure indoor or outdoor environments, but under time-varying wind conditions and coupled indoor and outdoor environments, the applicability is not clear. In this study, we proposed an improved adjoint probability method to identify the pollutant source location with time-varying inflows in street canyons and used scaled outdoor experiment data to verify the accuracy. The change in inflow velocity will affect the airflow structure inside the street canyons. Outdoor wind with a lower temperature will exchange heat with the air with a higher temperature inside the street canyon, taking away part of the heat and reducing the heat of the air inside the street canyons. Moreover, the room opening will produce some air disturbance, which is conducive to the heat exchange between the air near the opening and the outdoor wind. Furthermore, the fluctuations of the upper wind will influence the diffusion of the tracer gas. We conducted three cases to verify the accuracy of the source identification method. The results showed that the conditioned adjoint location probability (CALP) of each case was 0.06, 0.32, and 0.28. It implies that with limited pollutant information, the improved adjoint probability method can successfully identify the source location in the dynamic wind environments under coupled indoor and outdoor conditions.