General Model Research Articles

Generic Models for Predicting Coseismic Displacements of Earth Slopes Based on Numerical Analysis and Machine Learning Algorithm

Generic Models for Predicting Coseismic Displacements of Earth Slopes Based on Numerical Analysis and Machine Learning Algorithm

Predicting metro incident duration using structured data and unstructured text logs

Predicting metro incident duration is crucial for passengers and transit operators to choose appropriate response strategies. Most existing research focuses on structured data, the rich information embedded within unstructured incident logs is often neglected. This paper incorporates a probabilistic topic model tailored for short texts, the biterm topic model, into the generic incident duration prediction models. By capturing text co-occurrence patterns through Bayesian inference, the biterm topic model extracts hidden topics from incident narratives, and each topic serves as a condensed summary of detailed incident causes and countermeasures. These extracted topics are then combined with structured information to serve as predictors. We validated our model using five years of incident data from the Hong Kong Mass Transit Railway across two scenarios: with all incident information available and information revealed over time . The results demonstrate that our method significantly improves the prediction accuracy, particularly for incidents lasting longer than 30-minute.

Short-term voltage stability analysis in power systems: A voltage solvability indicator

This paper presents an analytical study for short-term voltage stability based on a novel indicator. First, it is shown that the reduced Jacobian matrix of network equations can be transformed to obtain an approximately symmetric matrix which is usually positive definite when the system under study is stable. The minimum eigenvalue of the matrix is suggested as a short-term Voltage Solvability Indicator (VSI). Then, properties of the reduced conductive matrix and susceptance matrix which have significant impacts on VSI are examined. The focus of the study is power systems with multiple constant power injections, while VSI can also be applied to analyze the impact of generic power system models. It is shown that as the penetration level of constant power injections increases, VSI decreases and the system becomes more vulnerable to voltage collapse. Finally, the relationship between the eigenvalues of the reduced Jacobian matrix and the structure-preserving Jacobian matrix is established. As a result, the suggested VSI can be efficiently computed by exploring the sparsity of structure-preserving Jacobian matrix of network equations. Case studies of several test systems including a simplified 575-machine 8117-bus East China power system are reported, demonstrating the effectiveness and practicability of the proposed method.

Redefining Neural Architecture Search of Heterogeneous Multinetwork Models by Characterizing Variation Operators and Model Components.

With neural architecture search (NAS) methods gaining ground on manually designed deep neural networks-even more rapidly as model sophistication escalates-the research trend is shifting toward arranging different and often increasingly complex NAS spaces. In this conjuncture, delineating algorithms which can efficiently explore these search spaces can result in a significant improvement over currently used methods, which, in general, randomly select the structural variation operator, hoping for a performance gain. In this article, we investigate the effect of different variation operators in a complex domain, that of multinetwork heterogeneous neural models. These models have an extensive and complex search space of structures as they require multiple subnetworks within the general model in order to answer different output types. From that investigation, we extract a set of general guidelines whose application is not limited to that particular type of model and are useful to determine the direction in which an architecture optimization method could find the largest improvement. To deduce the set of guidelines, we characterize both the variation operators, according to their effect on the complexity and performance of the model; and the models, relying on diverse metrics which estimate the quality of the different parts composing it.

Is Model-Based Science a Kind of Historical Science?

Abstract Philosophers have yet to provide a systematic analysis of the relationship between historical science and model-based science. In this paper I argue that prototypical model-based sciences exhibit features understood to be central to historical science. Philosophers of science have argued that historical scientists are distinctly concerned with inference to the best explanation, that explanations in historical science tend to increase in complexity over time, and that the explanations take the form of narratives. Using general circulation models in climate science as a reference, I illustrate how simulation models in model-based science share these features exhibited by historical science. That model based sciences share these features raises important philosophical questions about how we should understand prototypical types of scientific enquiry, including the relationship between experimental science, historical science, and model-based science. I conclude by exploring several options for how to accommodate the noted similarities within a more general taxonomy of the sciences.

A note on the stability of the Cauchy horizon in regular black holes without mass inflation

Unlike generic models of regular black holes (BHs) with nonzero surface gravity on both Cauchy and event horizons, an inner-degenerate counterpart with zero Cauchy horizon surface gravity was recently proposed. For this regular BH solution with spherical symmetry, we examine the stability of the Cauchy horizon from both classic and quantum mechanics viewpoints. We find that the classical perturbations do not lead to mass inflation at the Cauchy horizon, indicating classically stability, while quantum fluctuations cause it to be unstable quantum mechanically.

ECC-PolypDet: Enhanced CenterNet With Contrastive Learning for Automatic Polyp Detection.

Accurate polyp detection is critical for early colorectal cancer diagnosis. Although remarkable progress has been achieved in recent years, the complex colon environment and concealed polyps with unclear boundaries still pose severe challenges in this area. Existing methods either involve computationally expensive context aggregation or lack prior modeling of polyps, resulting in poor performance in challenging cases. In this paper, we propose the Enhanced CenterNet with Contrastive Learning (ECC-PolypDet), a two-stage training & end-to-end inference framework that leverages images and bounding box annotations to train a general model and fine-tune it based on the inference score to obtain a final robust model. Specifically, we conduct Box-assisted Contrastive Learning (BCL) during training to minimize the intra-class difference and maximize the inter-class difference between foreground polyps and backgrounds, enabling our model to capture concealed polyps. Moreover, to enhance the recognition of small polyps, we design the Semantic Flow-guided Feature Pyramid Network (SFFPN) to aggregate multi-scale features and the Heatmap Propagation (HP) module to boost the model's attention on polyp targets. In the fine-tuning stage, we introduce the IoU-guided Sample Re-weighting (ISR) mechanism to prioritize hard samples by adaptively adjusting the loss weight for each sample during fine-tuning. Extensive experiments on six large-scale colonoscopy datasets demonstrate the superiority of our model compared with previous state-of-the-art detectors.

Provably Trainable Rotationally Equivariant Quantum Machine Learning

Exploiting the power of quantum computation to realize superior machine learning algorithms has been a major research focus of recent years, but the prospects of quantum machine learning (QML) remain dampened by considerable technical challenges. A particularly significant issue is that generic QML models suffer from so-called barren plateaus in their training landscapes—large regions where cost function gradients vanish exponentially in the number of qubits employed, rendering large models effectively untrainable. A leading strategy for combating this effect is to build problem-specific models that take into account the symmetries of their data in order to focus on a smaller, relevant subset of Hilbert space. In this work, we introduce a family of rotationally equivariant QML models built upon the quantum Fourier transform, and leverage recent insights from the Lie-algebraic study of QML models to prove that (a subset of) our models do not exhibit barren plateaus. In addition to our analytical results we numerically test our rotationally equivariant models on a dataset of simulated scanning tunneling microscope images of phosphorus impurities in silicon, where rotational symmetry naturally arises, and find that they dramatically outperform their generic counterparts in practice. Published by the American Physical Society 2024

Bidirectional brain image translation using transfer learning from generic pre-trained models

Brain imaging plays a crucial role in the diagnosis and treatment of various neurological disorders, providing valuable insights into the structure and function of the brain. Techniques such as magnetic resonance imaging (MRI) and computed tomography (CT) enable non-invasive visualization of the brain, aiding in the understanding of brain anatomy, abnormalities, and functional connectivity. However, cost and radiation dose may limit the acquisition of specific image modalities, so medical image synthesis can be used to generate required medical images without actual addition. CycleGAN and other GANs are valuable tools for generating synthetic images across various fields. In the medical domain, where obtaining labeled medical images is labor-intensive and expensive, addressing data scarcity is a major challenge. Recent studies propose using transfer learning to overcome this issue. This involves adapting pre-trained CycleGAN models, initially trained on non-medical data, to generate realistic medical images. In this work, transfer learning was applied to the task of MR-CT image translation and vice versa using 18 pre-trained non-medical models, and the models were fine-tuned to have the best result. The models’ performance was evaluated using four widely used image quality metrics: Peak-signal-to-noise-ratio, Structural Similarity Index, Universal Quality Index, and Visual Information Fidelity. Quantitative evaluation and qualitative perceptual analysis by radiologists demonstrate the potential of transfer learning in medical imaging and the effectiveness of the generic pre-trained model. The results provide compelling evidence of the model’s exceptional performance, which can be attributed to the high quality and similarity of the training images to actual human brain images. These results underscore the significance of carefully selecting appropriate and representative training images to optimize performance in brain image analysis tasks.

Adding metabolic tasks to human GEM models to improve the study of gene targets and their associated toxicities

Genetic minimal cut sets (gMCS) are genes that must be deactivated simultaneously to avoid unwanted states in a metabolic model. The concept of gMCS can be applied to two different scenarios. First, it can be used to identify potential gene toxicities in generic or healthy cell models. Second, it can be used to develop genetic strategies to target cancer cells and prevent their proliferation. Up to now, gMCS have been evaluated using the traditional procedure of preventing biomass production. This paper proposes an additional way: using essential metabolic tasks, which any human cell should perform, to enlarge the set of unwanted states. Including this addition can significantly improve the study of toxicities and reveal targets that can be used to treat unhealthy cells. Excluding metabolic tasks can cause important information to be overlooked, which could impact the study’s success. Regarding toxicities, using the generic Human model, the number of detected generic toxicities with metabolic tasks increases from 106 to 281 (136 gMCSs of length 1 and 49 of length 2). We have used the following context-specific models to evaluate specific toxicities in different healthy tissues: blood, pancreas, liver, heart, and kidney. Again, considering metabolic tasks, we have found new toxicities (lengths 1 and 2) whose inactivation could damage these healthy tissues.Our research strategy has been applied to identify new cancer drug targets in two myeloma cell lines. We obtained new therapeutic targets of lengths 1 and 2 for each cell line. After analyzing the data, we conclude that incorporating metabolic tasks into cancer models can reveal important therapeutic targets previously disregarded by the conventional method of inhibiting biomass production. This approach also improves the evaluation of potential drug toxicities.

Understanding the public stigma of mental illness: a mixed-methods, multi-level, exploratory triangulation study

BackgroundThis study examines the role of themata in understanding mental health-related stigma. It is motivated by the need for alternative theoretical-methodological approaches beyond the dominant frameworks in education and contact-based anti-stigma public health efforts, which have shown mixed effects. Specifically, it addresses the need for a more nuanced framework in stigma research, one that is sensitive to the dialogues through which people relate themselves to mental health and stigma in context.MethodsThe research employs an exploratory mixed-methods approach, including the analysis of 529 news reports, 20 focus group discussions, and 19 one-to-one interviews, all concerning representations of shared living arrangements with someone perceived to have experiences of mental illness. Thematic analysis and natural language processing are used within a convergent triangulation design to analyze the data.ResultsWe found that mental health and illness were communicated through an overarching Self/Other thema and five subordinate themata: normal/abnormal, harm/non-harm, bounded/non-bounded, and moral/immoral. Despite familiarity with psychological distress and ‘modern’ explanations of mental illness, concerns about social identity motivated representations of mental illness as a predominantly permanent, negative form of personhood marked by abnormality, harm, distance, and immorality. Additionally, concerns about personal vulnerability, including historically rooted fears of contagion, motivated distancing representations of mental illness, rather than neutral portrayals.ConclusionsThemata have under-developed theoretical and methodological potential for addressing mental health-related stigma, particularly in their ability to describe the dynamic ways in which culture motivates people to both resist and reproduce stigma, partly through ambivalences, absences, tensions, and ambiguities in representation. A critical discussion is provided on how themata may support ecological strategies in mental health campaigns over generic models, emphasizing the need to understand group knowledge and contact dynamics to mitigate adverse effects. Themata Public Health Unintended Consequences Mixed Methods Behaviour Change Natural Language Processing.

Neutrino anarchy from flavor deconstruction

Flavor deconstruction refers to non-universal gauge extensions where the original gauge symmetry is deconstructed into separate copies, one for each family. A hierarchical chain of symmetry breaking provides an attractive low-scale solution to the flavor puzzle, consistent with flavor-changing neutral currents and finite naturalness. Although successful in explaining the origin of flavor hierarchies in the quark and charged lepton sectors, existing models have struggled with the large and seemingly anarchic mixing observed in neutrino oscillations. This letter identifies conditions under which neutrino anarchy may arise from flavor deconstruction in generic models with right-handed neutrinos. When deconstruction is applied to carefully chosen subgroups of the extended gauge symmetry, hierarchies in the seesaw formula cancel out.

Proactive Drive Failure Prediction for Cloud Storage System Through Semi-Supervised Learning

Proactive drive failure prediction can help operators handle the failing drives in advance, enhancing the storage system dependability. SSD and HDD failure prediction techniques are currently evolving towards a semi-supervised approach. In this paper, we are dedicated to enhancing the methodology for semi-supervised drive failure prediction from these aspects: design more powerful, robust yet generic models, mine drive failure modes and make the prediction model interpretable. Specifically, we propose two semi-supervised drive failure prediction models, DFP-VL and DFP-FL, from the perspective of reconstructing SMART data and learning SMART data's probability density, respectively. They capture the pattern of healthy drives, and failing drives can be detected when they deviate from the normal pattern. We mine two failure modes, slow deterioration and dramatic deterioration. Based on that, we design two failure detectors, ”ThresholdDetector” and ”HybridDetector”, to determine whether a drive deviates from the normal pattern. We evaluate the proposed methods on both SSD and HDD SMART data. DFP-VL is generic yet effective and can interpret the predicted results. DFP-FL has better performance than DFP-VL, but it cannot interpret the results. ThresholdDetector has low complexity and can detect most failing drives. HybridDetector has high complexity and can further improve the detection performance.

Windage loss characterisation for flywheel energy storage system: Model and experimental validation

In this paper, a windage loss characterisation strategy for Flywheel Energy Storage Systems (FESS) is presented. An effective windage loss modeling in FESS is essential for feasible and competitive design. Unlike generic aerodynamic loss models, FESS require particular attention to their unique characteristics i.e., vacuum, small airgaps, high angular speed, and presence of low friction rotor supports.The proposed model is based on several analytical and semi-empirical windage loss solutions for cylindrical and planar surface interactions, harmonised to manage the transition between laminar and turbulent flows. Also, the model is enriched by introducing corrections for rarefied gasses, using kinetic gas theory formulation. Therefore, it is possible the reduce the windage overestimation occurring with Navier–Stokes equation solutions for laminar flow. The model is compared to case studies from the literature featuring different boundary and operating conditions, to check consistency of all the harmonised models. A dedicated experimental test-rig is developed to validate the corrections to the model for rarefied gas at different vacuum levels. Then, a non-invasive characterisation approach for FESS windage losses in self-discharge phase is proposed and validated on the test-rig.

Multiview Clustering With Propagating Information Bottleneck.

In many practical applications, massive data are observed from multiple sources, each of which contains multiple cohesive views, called hierarchical multiview (HMV) data, such as image-text objects with different types of visual and textual features. Naturally, the inclusion of source and view relationships offers a comprehensive view of the input HMV data and achieves an informative and correct clustering result. However, most existing multiview clustering (MVC) methods can only process single-source data with multiple views or multisource data with single type of feature, failing to consider all the views across multiple sources. Observing the rich closely related multivariate (i.e., source and view) information and the potential dynamic information flow interacting among them, in this article, a general hierarchical information propagation model is first built to address the above challenging problem. It describes the process from optimal feature subspace learning (OFSL) of each source to final clustering structure learning (CSL). Then, a novel self-guided method named propagating information bottleneck (PIB) is proposed to realize the model. It works in a circulating propagation fashion, so that the resulting clustering structure obtained from the last iteration can "self-guide" the OFSL of each source, and the learned subspaces are in turn used to conduct the subsequent CSL. We theoretically analyze the relationship between the cluster structures learned in the CSL phase and the preservation of relevant information propagated from the OFSL phase. Finally, a two-step alternating optimization method is carefully designed for optimization. Experimental results on various datasets show the superiority of the proposed PIB method over several state-of-the-art methods.

Energy-dependent boosted dark matter from diffuse supernova neutrino background

Diffuse neutrinos from past supernovae in the Universe present us with a unique opportunity to test dark matter (DM) interactions. These neutrinos can scatter and boost the DM particles in the Milky Way halo to relativistic energies allowing us to detect them in terrestrial laboratories.Focusing on generic models of DM-neutrino and electron interactions, mediated by a vector or a scalar boson, we implement energy-dependent scattering cross-sections and perform detailed numerical analysis of DM attenuation due to electron scattering in-medium while propagating towards terrestrial experiments. We set new limits on DM-neutrino and electron interactions for DM with masses in the range ∼ (0.1, 104) MeV, using recent data from XENONnT, LUX-ZEPLIN, and PandaX-4T direct detection experiments. We demonstrate that consideration of energy-dependent cross-sections for DM interactions can significantly affect constraints previously derived under the assumption of constant cross-sections, modifying them by multiple orders of magnitude.

Understanding the impact of geotagging on location inference models for accurate generalization to non-geotagged datasets

For a location inference model to be successful, the properties of the geotagged tweets where the location inference model is developed and those of the non-geotagged tweets where the location inference model is applied need to match. We investigated location mentions within the tweet text field of 3,953,166 geotagged and 2,783,609 non-geotagged tweets across five of the most prominent Twitter sources. Specifically, we compared the frequency and the location entity types used to infer the locations within the two datasets. Overall we found statistically significant differences in location mentions between the two datasets. However, although statistically significant, thirteen of the fifteen analysed location entities, showed low effect sizes. We conclude that location inference models trained on geotagged datasets can generalize a non-geotagged dataset if special adjustments are made on the development of the location inference models.

Model-independent estimates for loop-induced baryon-number-violating nucleon decays

Baryon number is an accidental symmetry of the Standard Model (SM) Lagrangian that so far has been measured to be exactly preserved, although it is expected to be violated at higher energies. In this work we compute order-of-magnitude estimates for the matching contributions of generic ultraviolet models to effective operators that generate nucleon decay processes. This is done in a systematic and automated way using operators constructed from SM fields up to dimension nine and working in a framework that has proved useful in the study of lepton-number violation. For each of the operators we derive estimates for the rates of different nucleon-decay channels. These allow us to establish model-independent lower bounds on the underlying new-physics scale and identify potential correlations between the various decay modes. The results are most relevant for families of models that generate the considered operator. This analysis is especially timely given the expected future sensitivities in numerous experiments such as Hyper-K, DUNE, JUNO and THEIA.

The Impact of Using Dimpled Surfaces on Ahmed Body on The Flow Characteristics

Vehicle aerodynamics which directly affects fuel consumption values in vehicles has great importance in terms of vehicle stability, acceleration ability, vehicle noise, and environmental pollution. Ahmed car model is one of the most important generic models, so this popular car model is widely used in the literature on fluid mechanics and vehicle aerodynamics studies. This study investigates the effect of a passive drag reduction technique which is known as dimple surface, on Ahmed car model with a 35° slant angle. The flow control methods were applied numerically also these methods were compared with similar works. Analysis made in this study were conducted with 2.4 million cell numbers and the maximum drag reduction is 3.5 % with optimum dimple surfaces. Moreover, the applicability and manufacturability factors were also evaluated in terms of implementing the dimpling modification to today’s current road vehicles.

JT gravity from non-Abelian T-duality

We study the geometries obtained by performing super non-Abelian T-duality of the Principal Chiral Model on OSp(1|2). While the initial model represents an appropriate 3D supergravity background, interpretable as the superspace version of AdS3, the T-dual model fails solving the 3D supergravity torsion constraints. We argue that this has to do with a factorisation pattern taking place under dualisation: the dual 3D geometry can be rewritten as the supersymmetric version of AdS2, satisfying the supergravity constraints, fibered over what we interpret as the superspace equivalent of the standard bosonic line. We discuss an interesting connection between T-duals of generic Principal Chiral Models and Poisson sigma models. We exploit it to show that in a suitable limit the dual action studied in this work gives rise to JT (super)gravity.