Transonic buffet presents time-dependent aerodynamic characteristics associated with shock, turbulent boundary layer and their interactions. Despite strong nonlinearities and a large degree of freedom, there exists a dominant dynamic pattern of a buffet cycle, suggesting the low dimensionality of transonic buffet phenomena. This study seeks a low-dimensional representation of transonic airfoil buffet at a high Reynolds number with machine learning. Wall-modelled large-eddy simulations of flow over the OAT15A supercritical airfoil at two Mach numbers, $M_\infty = 0.715$ and 0.730, respectively producing non-buffet and buffet conditions, at a chord-based Reynolds number of ${Re} = 3\times 10^6$ are performed to generate the present datasets. We find that the low-dimensional nature of transonic airfoil buffet can be extracted as a sole three-dimensional latent representation through lift-augmented autoencoder compression. The current low-order representation not only describes the shock movement but also captures the moment when the separation occurs near the trailing edge in a low-order manner. We further show that it is possible to perform sensor-based reconstruction through the present low-dimensional expression while identifying the sensitivity with respect to aerodynamic responses. The present model trained at ${Re} = 3\times 10^6$ is lastly evaluated at the level of a real aircraft operation of ${Re} = 3\times 10^7$, exhibiting that the phase dynamics of lift is reasonably estimated from sparse sensors. The current study may provide a foundation towards data-driven real-time analysis of transonic buffet conditions under aircraft operation.

With ambitious action required to achieve global climate mitigation goals, climate change has become increasingly salient in the political arena. This article presents a dataset of climate change salience in 1,792 political manifestos of 620 political parties across different party families in forty-five OECD, European, and South American countries from 1990 to 2022. Importantly, our measure uniquely isolates climate change salience, avoiding the conflation with general environmental and sustainability content found in other work. Exploiting recent advances in supervised machine learning, we developed the dataset by fine-tuning a pre-trained multilingual transformer with human coding, employing a resource-efficient and replicable pipeline for multilingual text classification that can serve as a template for similar tasks. The dataset unlocks new avenues of research on the political discourse of climate change, on the role of parties in climate policy making, and on the political economy of climate change. We make the model and the dataset available to the research community.

The simulation of turbulent flow requires many degrees of freedom to resolve all the relevant time and length scales. However, due to the dissipative nature of the Navier–Stokes equations, the long-term dynamics is expected to lie on a finite-dimensional invariant manifold with fewer degrees of freedom. In this study, we build low-dimensional data-driven models of pressure-driven flow through a circular pipe. We impose the ‘shift-and-reflect’ symmetry to study the system in a minimal computational cell (e.g. the smallest domain size that sustains turbulence) at a Reynolds number of 2500. We build these models by using autoencoders to parametrise the manifold coordinates and neural ordinary differential equation to describe their time evolution. Direct numerical simulations (DNSs) typically require of the order of $\mathcal{O}(10^5)$ degrees of freedom, while our data-driven framework enables the construction of models with fewer than 20 degrees of freedom. Remarkably, these reduced-order models effectively capture crucial features of the flow, including the streak breakdown. In short-time tracking, these models accurately track the true trajectory for one Lyapunov time, as well as the leading Lyapunov exponent, while at long-times, they successfully capture key aspects of the dynamics such as Reynolds stresses and energy balance. The model can quantitatively capture key characteristics of the flow, including the streak breakdown and regeneration cycle. Additionally, we report new exact coherent states found in the DNS with the aid of these low-dimensional models. This approach leads to the discovery of seventeen previously unknown solutions within the turbulent pipe flow system, notably featuring relative periodic orbits characterised by the longest reported periods for such flow conditions.

With the growing amount of historical infrastructure data available to engineers, data-driven techniques have been increasingly employed to forecast infrastructure performance. In addition to algorithm selection, data preprocessing strategies for machine learning implementations plays an equally important role in ensuring accuracy and reliability. The present study focuses on pavement infrastructure and identifies four categories of strategies to preprocess data for training machine-learning-based forecasting models. The Long-Term Pavement Performance (LTPP) dataset is employed to benchmark these categories. Employing random forest as the machine learning algorithm, the comparative study examines the impact of data preprocessing strategies, the volume of historical data, and forecast horizon on the accuracy and reliability of performance forecasts. The strengths and limitations of each implementation strategy are summarized. Multiple pavement performance indicators are also analysed to assess the generalizability of the findings. Based on the results, several findings and recommendations are provided for short-to medium-term infrastructure management and decision-making: (i) in data-scarce scenarios, strategies that incorporate both explanatory variables and historical performance data provides better accuracy and reliability, (ii) to achieve accurate forecasts, the volume of historical data should at least span a time duration comparable to the intended forecast horizon, and (iii) for International Roughness Index and transverse crack length, a forecast horizon up to 5 years is generally achievable, but forecasts beyond a three-year horizon are not recommended for longitudinal crack length. These quantitative guidelines ultimately support more effective and reliable application of data-driven techniques in infrastructure performance forecasting.

Synthetic datasets, artificially generated to mimic real-world data while maintaining anonymization, have emerged as a promising technology in the financial sector, attracting support from regulators and market participants as a solution to data privacy and scarcity challenges limiting machine learning (ML) deployment. This article argues that synthetic data’s effects on financial markets depend critically on how these technologies are embedded within existing ML infrastructural ‘stacks’ rather than on their intrinsic properties. We identify three key tensions that will determine whether adoption proves beneficial or harmful: (1) data circulability versus opacity, particularly the ‘double opacity’ problem arising from stacked ML systems, (2) model-induced scattering versus model-induced herding in market participant behavior, and (3) flattening versus deepening of data platform power. These tensions directly correspond to core regulatory priorities around model risk management, systemic risk, and competition policy. Using financial audit as a case study, we demonstrate how these tensions interact in practice and propose governance frameworks, including a synthetic data labeling regime to preserve contextual information when datasets cross organizational boundaries.

A deep-learning-based closure model to address energy loss in low-dimensional surrogate models based on proper-orthogonal-decomposition (POD) modes is introduced. Using a transformer-encoder block with an easy-attention mechanism, the model predicts the spatial probability density function of fluctuations not captured by the truncated POD modes. The methodology is demonstrated on the wake of the Windsor body at yaw angles of $\delta = [2.5^\circ ,5^\circ ,7.5^\circ ,10^\circ ,12.5^\circ ]$, with $\delta = 7.5^\circ$ as a test case, and in a realistic urban environment at wind directions of $\delta = [-45^\circ ,-22.5^\circ ,0^\circ ,22.5^\circ ,45^\circ ]$, with $\delta = 0^\circ$ as a test case. Key coherent modes are identified by clustering them based on dominant frequency dynamics using Hotelling’s $T^2$ on the spectral properties of temporal coefficients. These coherent modes account for nearly $60 \,\%$ and $75 \,\%$ of the total energy for the Windsor body and the urban environment, respectively. For each case, a common POD basis is created by concatenating coherent modes from training angles and orthonormalising the set without losing information. Transformers with different size on the attention layer, (64, 128 and 256), are trained to model the missing fluctuations in the Windsor body case. Larger attention sizes always improve predictions for the training set, but the transformer with an attention layer of size 256 slightly overshoots the fluctuation predictions in the Windsor body test set because they have lower intensity than in the training cases. A single transformer with an attention size of 256 is trained for the urban flow. In both cases, adding the predicted fluctuations close the energy gap between the reconstruction and the original flow field, improving predictions for energy, root-mean-square velocity fluctuations and instantaneous flow fields. For instance, in the Windsor body case, the deepest architecture reduces the mean energy error from $37 \,\%$ to $12 \,\%$ and decreases the Kullback–Leibler divergence of velocity distributions from ${\mathcal{D}}_{\mathcal{KL}}=0.2$ to below ${\mathcal{D}}_{\mathcal{KL}}=0.026$.

Politicians’ presentation of self is central to election efforts. For these efforts to be successful, they need voters to receive and believe the messages they communicate. We examine the relationship between politicians’ communications and voters’ perceptions of their ideology. Using the content of politicians’ ideological presentation of self through social media communications, we create a measure of messaging ideology for all congressional candidates between 2018 and 2022 and all congressional officeholders between 2012 and 2022 along with voter perceptions of candidate ideology during the same time period. Using these measures, our work shows voters’ perceptions of candidate ideology are strongly related to messaging, even after controlling for incumbent voting behavior. We also examine how the relationship between politician messaging and voter perceptions changes relative to other information about the politician and in different electoral contexts. On the whole, voters’ perceptions of candidate ideology are strongly correlated with politician communications.

Despite their widespread use, purely data-driven methods often suffer from overfitting, lack of physical consistency, and high data dependency, particularly when physical constraints are not incorporated. This study introduces a novel data assimilation approach that integrates Graph Neural Networks (GNNs) with optimization techniques to enhance the accuracy of mean flow reconstruction, using Reynolds-averaged Navier–Stokes (RANS) equations as a baseline. The method leverages the adjoint approach, incorporating RANS-derived gradients as optimization terms during GNN training, ensuring that the learned model adheres to physical laws and maintains consistency. Additionally, the GNN framework is well-suited for handling unstructured data, which is common in the complex geometries encountered in computational fluid dynamics. The GNN is interfaced with the finite element method for numerical simulations, enabling accurate modeling in unstructured domains. We consider the reconstruction of mean flow past bluff bodies at low Reynolds numbers as a test case, addressing tasks such as sparse data recovery, denoising, and inpainting of missing flow data. The key strengths of the approach lie in its integration of physical constraints into the GNN training process, leading to accurate predictions with limited data, making it particularly valuable when data are scarce or corrupted. Results demonstrate significant improvements in the accuracy of mean flow reconstructions, even with limited training data, compared to analogous purely data-driven models.

Within the context of machine learning-based closure mappings for Reynolds-averaged Navier Stokes turbulence modelling, physical realisability is often enforced using ad hoc postprocessing of the predicted anisotropy tensor. In this study, we address the realisability issue via a new physics-based loss function that penalises non-realisable results during training, thereby embedding a preference for realisable predictions into the model. Additionally, we propose a new framework for data-driven turbulence modelling which retains the stability and conditioning of optimal eddy viscosity-based approaches while embedding equivariance. Several modifications to the tensor basis neural network to enhance training and testing stability are proposed. We demonstrate the conditioning, stability and generalisation of the new framework and model architecture on three flows: flow over a flat plate, flow over periodic hills and flow through a square duct. The realisability-informed loss function is demonstrated to significantly increase the number of realisable predictions made by the model when generalising to a new flow configuration. Altogether, the proposed framework enables the training of stable and equivariant anisotropy mappings, with more physically realisable predictions on new data. We make our code available for use and modification by others. Moreover, as part of this study, we explore the applicability of Kolmogorov–Arnold networks to turbulence modelling, assessing its potential to address nonlinear mappings in the anisotropy tensor predictions and demonstrating promising results for the flat plate case.

This research paper aimed to develop a supervised machine learning (ML) approach that learns and predicts data-based culling from farm information that reflects the criteria of the decisions taken to cull a cow by a farm manager. Data containing the features of milk yield, days in milk, lactation number, pregnancy status, days open and days pregnant were obtained from January to December 2020 from dairy cows on a large dairy farm in northern Mexico. The cows were labelled as those that were data-based culled (Cull) and those that were not culled (Stay). Six supervised ML algorithms were evaluated in a binary classification including logistic regression (LR), Gaussian naïve Bayes (GNB), k-nearest neighbors (k-NN), support vector machine (SVM), random forest (RF) and multilayer perceptron (MLP). Each model was subjected to hyperparameter optimization using a grid search approach combined with tenfold stratified cross-validation. This ensured that the class imbalance (Cull vs. Stay) was accounted during model evaluation. The best-performing model for each algorithm was selected on cross-validated accuracy. To evaluate the prediction performance of the ML algorithms on both labels from learned data, the metrics accuracy, precision, recall, F1-score and the Matthews correlation coefficient (MCC) were employed. Accuracy among all classifiers was >0.90. The poorest prediction performance was observed in GNB (MCC = 0.50) and LR (MCC = 0.72). Conversely, the rest of the classifiers achieved superior prediction performance in learning the specific culling criteria, reaching an MCC score >0.91. Overall, culling criteria can be learned and predicted by ML algorithms and their performance varies among classifiers. This study identified RF as the best performing algorithm, but k-NN, SVM and MLP are possible candidates to be used in on-farm conditions. To increase their reliability, these approaches need to be tested in several farms, under different scenarios and varieties of features.