AbstractObjectiveTo systematically evaluate which lesion‐based imaging features and methods allow for the best statistical prediction of poststroke deficits across independent datasets.MethodsWe utilized imaging and clinical data from three independent datasets of patients experiencing acute stroke (N1 = 109, N2 = 638, N3 = 794) to statistically predict acute stroke severity (NIHSS) based on lesion volume, lesion location, and structural and functional disconnection with the lesion location using normative connectomes.ResultsWe found that prediction models trained on small single‐center datasets could perform well using within‐dataset cross‐validation, but results did not generalize to independent datasets (median R2N1 = 0.2%). Performance across independent datasets improved using large single‐center training data (R2N2 = 15.8%) and improved further using multicenter training data (R2N3 = 24.4%). These results were consistent across lesion attributes and prediction models. Including either structural or functional disconnection in the models outperformed prediction based on volume or location alone (P < 0.001, FDR‐corrected).InterpretationWe conclude that (1) prediction performance in independent datasets of patients with acute stroke cannot be inferred from cross‐validated results within a dataset, as performance results obtained via these two methods differed consistently, (2) prediction performance can be improved by training on large and, importantly, multicenter datasets, and (3) structural and functional disconnection allow for improved prediction of acute stroke severity.

Deep learning has allowed for remarkable progress in many medical scenarios. Deep learning prediction models often require 105-107 examples. It is currently unknown whether deep learning can also enhance predictions of symptoms post-stroke in real-world samples of stroke patients that are often several magnitudes smaller. Such stroke outcome predictions however could be particularly instrumental in guiding acute clinical and rehabilitation care decisions. We here compared the capacities of classically used linear and novel deep learning algorithms in their prediction of stroke severity. Our analyses relied on a total of 1430 patients assembled from the MRI-Genetics Interface Exploration collaboration and a Massachusetts General Hospital-based study. The outcome of interest was National Institutes of Health Stroke Scale-based stroke severity in the acute phase after ischaemic stroke onset, which we predict by means of MRI-derived lesion location. We automatically derived lesion segmentations from diffusion-weighted clinical MRI scans, performed spatial normalization and included a principal component analysis step, retaining 95% of the variance of the original data. We then repeatedly separated a train, validation and test set to investigate the effects of sample size; we subsampled the train set to 100, 300 and 900 and trained the algorithms to predict the stroke severity score for each sample size with regularized linear regression and an eight-layered neural network. We selected hyperparameters on the validation set. We evaluated model performance based on the explained variance (R2) in the test set. While linear regression performed significantly better for a sample size of 100 patients, deep learning started to significantly outperform linear regression when trained on 900 patients. Average prediction performance improved by ∼20% when increasing the sample size 9× [maximum for 100 patients: 0.279 ± 0.005 (R2, 95% confidence interval), 900 patients: 0.337 ± 0.006]. In summary, for sample sizes of 900 patients, deep learning showed a higher prediction performance than typically employed linear methods. These findings suggest the existence of non-linear relationships between lesion location and stroke severity that can be utilized for an improved prediction performance for larger sample sizes.

We introduce GAUCHE, an open-source library for GAUssian processes in CHEmistry. Gaussian processes have long been a cornerstone of probabilistic machine learning, affording particular advantages for uncertainty quantification and Bayesian optimisation. Extending Gaussian processes to molecular representations, however, necessitates kernels defined over structured inputs such as graphs, strings and bit vectors. By providing such kernels in a modular, robust and easy-to-use framework, we seek to enable expert chemists and materials scientists to make use of state-of-the-art black-box optimization techniques. Motivated by scenarios frequently encountered in practice, we showcase applications for GAUCHE in molecular discovery, chemical reaction optimisation and protein design. The codebase is made available at https://github.com/leojklarner/gauche.

We applied computational style transfer, specifically coloration and brush stroke style, to achromatic images of a ghost painting beneath Vincent van Gogh's Still life with meadow flowers and roses. Our method is an extension of our previous work in that it used representative artworks by the ghost painting\rq s author to train a Generalized Adversarial Network (GAN) for integrating styles learned from stylistically distinct groups of works. An effective amalgam of these learned styles is then transferred to the target achromatic work.

Introduction: Predicting individual outcomes post-stroke with the highest possible accuracy is a crucial steppingstone in the realization of precision medicine. We here evaluated various types of lesion information in their capacity to predict stroke severity in a large cohort of patients with acute ischemic stroke. Methods: A total of 1,075 patients of the MRI-GENIE study [age: 64.2(14.7), 38% women] contributed to analyses (N=792 as train and N=283 as test sample). We employed ridge regression with hyperparameter optimization and nested five-fold cross-validation to predict acute NIHSS-based stroke severity. Our baseline model considered DWI lesion volume. Further models tested structural lesion location, indirect structural and functional lesion connectivity, individually and combined with lesion volume. Data was preprocessed by principal component analysis (PCA), retaining 95% of the variance of the original data. Model performance was compared in terms of explained variance (R-squared) and 95% confidence intervals in the outer loop of the training set and the test set. Results: Structural lesion connectivity enabled the highest prediction performance in the test set ( Figure 1 ). Prediction performance did not change notably with inclusion of lesion volume information. This contrasted with the combination of functional lesion connectivity and lesion volume that achieved a substantially higher prediction performance compared to functional lesion connectivity or lesion volume in isolation. Lesion location resulted in a prediction performance that was in the range of lesion volume alone and significantly lower than the ones of structural or functional connectivity with lesion volume. Conclusions: Structural connectivity facilitated the most convincing prediction of stroke severity. The capacity of functional lesion connectivity was substantially improved by the inclusion of lesion volume, indicating both represent complementary information.

We discuss the problem of computationally generating images resembling those of lost cultural patrimony, specifically two-dimensional artworks such as paintings and drawings. We view the problem as one of computing an estimate of the image in the lost work that best conforms to surviving information in a variety of forms: works by the source artist, including preparatory works such as cartoons for the target work; copies of the target by other artists; other works by these artists that reveal aspects of their style; historical knowledge of art methods. and materials; stylistic conventions of the relevant era; textual descriptions of the lost work and as well as more generally, images associated with stories given by the target's title. Some of the general information linking images and text can be learned from large corpora of natural photographs and accompanying text scraped from the web. We present some preliminary, proof-of-concept simulations for recovering lost artworks with a special focus on textual information about target artworks. We outline our future directions, such as methods for assessing the contributions of different forms of information in the overall task of recovering lost artworks.

We present a novel bi-modal system based on deep networks to address the problem of learning associations and simple meanings of objects depicted in"authored"images, such as fine art paintings and drawings. Our overall system processes both the images and associated texts in order to learn associations between images of individual objects, their identities and the abstract meanings they signify. Unlike past deep nets that describe depicted objects and infer predicates, our system identifies meaning-bearing objects ("signifiers") and their associations ("signifieds") as well as basic overall meanings for target artworks. Our system had precision of 48% and recall of 78% with an F1 metric of 0.6 on a curated set of Dutch vanitas paintings, a genre celebrated for its concentration on conveying a meaning of great import at the time of their execution. We developed and tested our system on fine art paintings but our general methods can be applied to other authored images.

Photoswitchable molecules display two or more isomeric forms that may be accessed using light. Separating the electronic absorption bands of these isomers is key to selectively addressing a specific isomer and achieving high photostationary states whilst overall red-shifting the absorption bands serves to limit material damage due to UV-exposure and increases penetration depth in photopharmacological applications. Engineering these properties into a system through synthetic design however, remains a challenge. Here, we present a data-driven discovery pipeline for molecular photoswitches underpinned by dataset curation and multitask learning with Gaussian processes. In the prediction of electronic transition wavelengths, we demonstrate that a multioutput Gaussian process (MOGP) trained using labels from four photoswitch transition wavelengths yields the strongest predictive performance relative to single-task models as well as operationally outperforming time-dependent density functional theory (TD-DFT) in terms of the wall-clock time for prediction. We validate our proposed approach experimentally by screening a library of commercially available photoswitchable molecules. Through this screen, we identified several motifs that displayed separated electronic absorption bands of their isomers, exhibited red-shifted absorptions, and are suited for information transfer and photopharmacological applications. Our curated dataset, code, as well as all models are made available at https://github.com/Ryan-Rhys/The-Photoswitch-Dataset

Models of human motion commonly focus either on trajectory prediction or action classification but rarely both. The marked heterogeneity and intricate compositionality of human motion render each task vulnerable to the data degradation and distributional shift common to real-world scenarios. A sufficiently expressive generative model of action could in theory enable data conditioning and distributional resilience within a unified framework applicable to both tasks. Here we propose a novel architecture based on hierarchical variational autoencoders and deep graph convolutional neural networks for generating a holistic model of action over multiple time-scales. We show this Hierarchical Graph-convolutional Variational Autoencoder (HGVAE) to be capable of generating coherent actions, detecting out-of-distribution data, and imputing missing data by gradient ascent on the model’s posterior. Trained and evaluated on H3.6M and the largest collection of open source human motion data, AMASS, we show HG-VAE can facilitate downstream discriminative learning better than baseline models.

The optical and UV variability of the majority of active galactic nuclei may be related to the reprocessing of rapidly changing X-ray emission from a more compact region near the central black hole. Such a reprocessing model would be characterized by lags between X-ray and optical/UV emission due to differences in light travel time. Observationally, however, such lag features have been difficult to detect due to gaps in the lightcurves introduced through factors such as source visibility or limited telescope time. In this work, Gaussian process regression is employed to interpolate the gaps in the Swift X-ray and UV lightcurves of the narrow-line Seyfert 1 galaxy Mrk 335. In a simulation study of five commonly employed analytic Gaussian process kernels, we conclude that the Matern 1/2 and rational quadratic kernels yield the most well-specified models for the X-ray and UVW2 bands of Mrk 335. In analyzing the structure functions of the Gaussian process lightcurves, we obtain a broken power law with a break point at 125 days in the UVW2 band. In the X-ray band, the structure function of the Gaussian process lightcurve is consistent with a power law in the case of the rational quadratic kernel while a broken power law with a break point at 66 days is obtained from the Matern 1/2 kernel. The subsequent cross-correlation analysis is consistent with previous studies and furthermore shows tentative evidence for a broad X-ray-UV lag feature of up to 30 days in the lag-frequency spectrum where the significance of the lag depends on the choice of Gaussian process kernel.

We describe the application of convolutional neural network style transfer to the problem of improved visualization of underdrawings and ghost-paintings in fine art oil paintings. Such underdrawings and hidden paintings are typically revealed by x-ray or infrared techniques which yield images that are grayscale, and thus devoid of color and full style information. Past methods for inferring color in underdrawings have been based on physical x-ray fluorescence spectral imaging of pigments in ghost-paintings and are thus expensive, time consuming, and require equipment not available in most conservation studios. Our algorithmic methods do not need such expensive physical imaging devices. Our proof-of-concept system, applied to works by Pablo Picasso and Leonardo, reveal colors and designs that respect the natural segmentation in the ghost-painting. We believe the computed images provide insight into the artist and associated oeuvre not available by other means. Our results strongly suggest that future applications based on larger corpora of paintings for training will display color schemes and designs that even more closely resemble works of the artist. For these reasons refinements to our methods should find wide use in art conservation, connoisseurship, and art analysis.

We apply generative adversarial convolutional neural networks to the problem of style transfer to underdrawings and ghost-images in x-rays of fine art paintings with a special focus on enhancing their spatial resolution. We build upon a neural architecture developed for the related problem of synthesizing high-resolution photo-realistic image from semantic label maps. Our neural architecture achieves high resolution through a hierarchy of generators and discriminator sub-networks, working throughout a range of spatial resolutions. This coarse-to-fine generator architecture can increase the effective resolution by a factor of eight in each spatial direction, or an overall increase in number of pixels by a factor of 64. We also show that even just a few examples of human-generated image segmentations can greatly improve -- qualitatively and quantitatively -- the generated images. We demonstrate our method on works such as Leonardo's Madonna of the carnation and the underdrawing in his Virgin of the rocks, which pose several special problems in style transfer, including the paucity of representative works from which to learn and transfer style information.

Abstract The task of predicting human motion is complicated by the natural heterogeneity and compositionality of actions, necessitating robustness to distributional shifts as far as out‐of‐distribution (OoD). Here, we formulate a new OoD benchmark based on the Human3.6M and Carnegie Mellon University (CMU) motion capture datasets, and introduce a hybrid framework for hardening discriminative architectures to OoD failure by augmenting them with a generative model. When applied to current state‐of‐the‐art discriminative models, we show that the proposed approach improves OoD robustness without sacrificing in‐distribution performance, and can theoretically facilitate model interpretability. We suggest human motion predictors ought to be constructed with OoD challenges in mind, and provide an extensible general framework for hardening diverse discriminative architectures to extreme distributional shift. The code is available at: https://github.com/bouracha/OoDMotion.

The automatic analysis of fine art paintings presents a number of novel technical challenges to artificial intelligence, computer vision, machine learning, and knowledge representation quite distinct from those arising in the analysis of traditional photographs. The most important difference is that many realist paintings depict stories or episodes in order to convey a lesson, moral, or meaning. One early step in automatic interpretation and extraction of meaning in artworks is the identifications of figures (“actors”). In Christian art, specifically, one must identify the actors in order to identify the Biblical episode or story depicted, an important step in “understanding” the artwork. We designed an auto-matic system based on deep convolutional neural net-works and simple knowledge database to identify saints throughout six centuries of Christian art based in large part upon saints’ symbols or attributes. Our work rep-resents initial steps in the broad task of automatic se- mantic interpretation of messages and meaning in fine art.

The space of synthesizable molecules is greater than $10^{60}$, meaning only a vanishingly small fraction of these molecules have ever been realized in the lab. In order to prioritize which regions of this space to explore next, synthetic chemists need access to accurate molecular property predictions. While great advances in molecular machine learning have been made, there is a dearth of benchmarks featuring properties that are useful for the synthetic chemist. Focussing directly on the needs of the synthetic chemist, we introduce the Photoswitch Dataset, a new benchmark for molecular machine learning where improvements in model performance can be immediately observed in the throughput of promising molecules synthesized in the lab. Photoswitches are a versatile class of molecule for medical and renewable energy applications where a molecule's efficacy is governed by its electronic transition wavelengths. We demonstrate superior performance in predicting these wavelengths compared to both time-dependent density functional theory (TD-DFT), the incumbent first principles quantum mechanical approach, as well as a panel of human experts. Our baseline models are currently being deployed in the lab as part of the decision process for candidate synthesis. It is our hope that this benchmark can drive real discoveries in photoswitch chemistry and that future benchmarks can be introduced to pivot learning algorithm development to benefit more expansive areas of synthetic chemistry.

Animal behaviour is complex and the amount of data in the form of video, if extracted, is copious. Manual analysis of behaviour is massively limited by two insurmountable obstacles, the complexity of the behavioural patterns and human bias. Automated visual analysis has the potential to eliminate both of these issues and also enable continuous analysis allowing a much higher bandwidth of data collection which is vital to capture complex behaviour at many different time scales. Behaviour is not confined to a finite set modules and thus we can only model it by inferring the generative distribution. In this way unpredictable, anomalous behaviour may be considered. Here we present a method of unsupervised behavioural analysis from nothing but high definition video recordings taken from a single, fixed perspective. We demonstrate that the identification of stereotyped rodent behaviour can be extracted in this way.

Neural style transfer, first proposed by Gatys et al. (2015), can be used to create novel artistic work through rendering a content image in the form of a style image. We present a novel method of reconstructing lost artwork, by applying neural style transfer to x-radiographs of artwork with secondary interior artwork beneath a primary exterior, so as to reconstruct lost artwork. Finally we reflect on AI art exhibitions and discuss the social, cultural, ethical, and philosophical impact of these technical innovations.

This report describes the design, implementation and analysis of a non-blocking halo exchange routine as an alternative to the blocking halo exchange routine in the lattice Boltzmann code Ludwig. The alternative, non-blocking, routine is implemented in such a way to allow work-communication overlap. Detailed benchmarks in this report show that the non-blocking version is a good alternative even without any work-communication overlap. Work-Communication overlap can be used to improve the performance of the non-blocking routine. Development and benchmarking were conducted on the UK national supercomputer, ARCHER.

This report describes the design, implementation and analysis of a non-blocking halo exchange routine as an alternative to the blocking halo exchange routine in the lattice Boltzmann code Ludwig. The alternative non-blocking routine is implemented in such a way to allow workcommunication overlap. Detailed benchmarks in this report show that the non-blocking version is a good alternative with some advantages and no significant disadvantages to the blocking version even without any work-communication overlap. Work-Communication overlap can be used to improve the performance of the non-blocking routine. Development and benchmarking were conducted on the UK national supercomputer, ARCHER. Chapter

We perform Monte Carlo simulations on the 2-dimensional Ising model in a zero magnetic field, using the Metropolis and Wolff Cluster algorithms to obtain computational results for thermodynamic quantities and the dynamic critical exponents of the algorithms. The Ising model in 2-dimensions and zero magnetic field undergoes a second order phase transition from a phase with no magnetisation to one with spontaneous magnetisation. This phase transition is modelled by our simulations with the behaviour of magnetisation, energy, magnetic susceptibility and heat capacity at phase transition all being accurately valued. We used finite size scaling of the lattice at critical temperature to derive the value of the dynamic critical exponent for both algorithms of z = 2.11 ± 0.073 for Metropolis and 0.47 ± 0.051 for Wolff Cluster which compares well with accepted quantities of zmetropolis = 2.17 and zcluster = 0.52. Critical exponents of derived observables magnetic susceptibility (γ) and heat capacity (α) were found to be γ = 1.76 ± 0.023 and α = 0.29 ± 1.5. γ compared well with the accepted value of γ = 7/4. The discrepancy of α from the accepted value 0 was discussed. Chapter