Data augmentation is an effective technique for improving the accuracy of modern image classifiers. However, current data augmentation implementations are manually designed. In this paper, we ...describe a simple procedure called AutoAugment to automatically search for improved data augmentation policies. In our implementation, we have designed a search space where a policy consists of many sub-policies, one of which is randomly chosen for each image in each mini-batch. A sub-policy consists of two operations, each operation being an image processing function such as translation, rotation, or shearing, and the probabilities and magnitudes with which the functions are applied. We use a search algorithm to find the best policy such that the neural network yields the highest validation accuracy on a target dataset. Our method achieves state-of-the-art accuracy on CIFAR-10, CIFAR-100, SVHN, and ImageNet (without additional data). On ImageNet, we attain a Top-1 accuracy of 83.5% which is 0.4% better than the previous record of 83.1%. On CIFAR-10, we achieve an error rate of 1.5%, which is 0.6% better than the previous state-of-the-art. Augmentation policies we find are transferable between datasets. The policy learned on ImageNet transfers well to achieve significant improvements on other datasets, such as Oxford Flowers, Caltech-101, Oxford-IIT Pets, FGVC Aircraft, and Stanford Cars.
Making kirigami-inspired cuts into a sheet has been shown to be an effective way of designing stretchable materials with metamorphic properties where the 2D shape can transform into complex 3D ...shapes. However, finding the optimal solutions is not straightforward as the number of possible cutting patterns grows exponentially with system size. Here, we report on how machine learning (ML) can be used to approximate the target properties, such as yield stress and yield strain, as a function of cutting pattern. Our approach enables the rapid discovery of kirigami designs that yield extreme stretchability as verified by molecular dynamics (MD) simulations. We find that convolutional neural networks, commonly used for classification in vision tasks, can be applied for regression to achieve an accuracy close to the precision of the MD simulations. This approach can then be used to search for optimal designs that maximize elastic stretchability with only 1000 training samples in a large design space of ∼4×10^{6} candidate designs. This example demonstrates the power and potential of ML in finding optimal kirigami designs at a fraction of iterations that would be required of a purely MD or experiment-based approach, where no prior knowledge of the governing physics is known or available.
Including prior knowledge is important for effective machine learning models in physics and is usually achieved by explicitly adding loss terms or constraints on model architectures. Prior knowledge ...embedded in the physics computation itself rarely draws attention. We show that solving the Kohn-Sham equations when training neural networks for the exchange-correlation functional provides an implicit regularization that greatly improves generalization. Two separations suffice for learning the entire one-dimensional H_{2} dissociation curve within chemical accuracy, including the strongly correlated region. Our models also generalize to unseen types of molecules and overcome self-interaction error.
We discover many new crystalline solid materials with fast single crystal Li ion conductivity at room temperature, discovered through density functional theory simulations guided by machine ...learning-based methods. The discovery of new solid Li superionic conductors is of critical importance to the development of safe all-solid-state Li-ion batteries. With a predictive universal structure–property relationship for fast ion conduction not well understood, the search for new solid Li ion conductors has relied largely on trial-and-error computational and experimental searches over the last several decades. In this work, we perform a guided search of materials space with a machine learning (ML)-based prediction model for material selection and density functional theory molecular dynamics (DFT-MD) simulations for calculating ionic conductivity. These materials are screened from over 12 000 experimentally synthesized and characterized candidates with very diverse structures and compositions. When compared to a random search of materials space, we find that the ML-guided search is 2.7 times more likely to identify fast Li ion conductors, with at least a 44 times improvement in the log-average of room temperature Li ion conductivity. The F1 score of the ML-based model is 0.50, 3.5 times better than the F1 score expected from completely random guesswork. In a head-to-head competition against six Ph.D. students working in the field, we find that the ML-based model doubles the F1 score of human experts in its ability to identify fast Li-ion conductors from atomistic structure with a 1000-fold increase in speed, clearly demonstrating the utility of this model for the research community. In addition to having high predicted Li-ion conductivity, all materials reported here lack transition metals to enhance stability against reduction by the Li metal anode and are predicted to exhibit low electronic conduction, high stability against oxidation, and high thermodynamic stability, making them promising candidates for solid-state electrolyte applications on these several essential metrics.
Modern lithium ion batteries are often desired to operate at a wide electrochemical window to maximize energy densities. While pushing the limit of cutoff potentials allows batteries to provide ...greater energy densities with enhanced specific capacities and higher voltage outputs, it raises key challenges with thermodynamic and kinetic stability in the battery. This is especially true for layered lithium transition-metal oxides, where capacities can improve but stabilities are compromised as wider electrochemical windows are applied. To overcome the above-mentioned challenges, we used atomic layer deposition to develop a LiAlF4 solid thin film with robust stability and satisfactory ion conductivity, which is superior to commonly used LiF and AlF3. With a predicted stable electrochemical window of approximately 2.0 ± 0.9 to 5.7 ± 0.7 V vs Li+/Li for LiAlF4, excellent stability was achieved for high Ni content LiNi0.8Mn0.1Co0.1O2 electrodes with LiAlF4 interfacial layer at a wide electrochemical window of 2.75–4.50 V vs Li+/Li.
Building instance segmentation models that are data-efficient and can handle rare object categories is an important challenge in computer vision. Leveraging data augmentations is a promising ...direction towards addressing this challenge. Here, we perform a systematic study of the Copy-Paste augmentation (e.g., 13, 12) for instance segmentation where we randomly paste objects onto an image. Prior studies on Copy-Paste relied on modeling the surrounding visual context for pasting the objects. However, we find that the simple mechanism of pasting objects randomly is good enough and can provide solid gains on top of strong baselines. Furthermore, we show Copy-Paste is additive with semi-supervised methods that leverage extra data through pseudo labeling (e.g. self-training). On COCO instance segmentation, we achieve 49.1 mask AP and 57.3 box AP, an improvement of +0.6 mask AP and +1.5 box AP over the previous state-of-the-art. We further demonstrate that Copy-Paste can lead to significant improvements on the LVIS benchmark. Our baseline model outperforms the LVIS 2020 Challenge winning entry by +3.6 mask AP on rare categories. 1
The dynamical glass transition is typically taken to be the temperature at which a glassy liquid is no longer able to equilibrate on experimental timescales. Consequently, the physical properties of ...these systems just above or below the dynamical glass transition, such as viscosity, can change by many orders of magnitude over long periods of time following external perturbation. During this progress toward equilibrium, glassy systems exhibit a history dependence that has complicated their study. In previous work, we bridged the gap between structure and dynamics in glassy liquids above their dynamical glass transition temperatures by introducing a scalar field called “softness,” a quantity obtained using machine-learning methods. Softness is designed to capture the hidden patterns in relative particle positions that correlate strongly with dynamical rearrangements of particle positions. Here we show that the out-of-equilibrium behavior of a model glass-forming system can be understood in terms of softness. To do this we first demonstrate that the evolution of behavior following a temperature quench is a primarily structural phenomenon: The structure changes considerably, but the relationship between structure and dynamics remains invariant. We then show that the relaxation time can be robustly computed from structure as quantified by softness, with the same relation holding both in equilibrium and as the system ages. Together, these results show that the history dependence of the relaxation time in glasses requires knowledge only of the softness in addition to the usual state variables.
Emerging applications of metal–organic frameworks (MOFs) in electronic devices will benefit from the design and synthesis of intrinsically, highly electronically conductive MOFs. However, very few ...are known to exist. It is a challenging task to search for electronically conductive MOFs within the tens of thousands of reported MOF structures. Using a new strategy (i.e., transfer learning) of combining machine learning techniques, statistical multivoting, and ab initio calculations, we screened 2932 MOFs and identified 6 MOF crystal structures that are metallic at the level of semilocal DFT band theory: Mn2Re6X8(CN)64 (X = S, Se,Te), MnRe3Te4(CN)3, HgSCN4CoNCS4, and CdC4. Five of these structures have been synthesized and reported in the literature, but their electrical characterization has not been reported. Our work demonstrates the potential power of machine learning in materials science to aid in down-selecting from large numbers of potential candidates and provides the information and guidance to accelerate the discovery of novel advanced materials.
Determining structural transformations in amorphous solids is challenging due to the paucity of structural signatures. The effect of the transitions on the properties of the solid can be significant ...and important for applications. Moreover, such transitions may not be discernible in the behavior of the total energy or the volume of the solid as a function of the variables that identify its phases. These issues arise in the context of lithiation of amorphous silicon (a-Si), a promising anode material for high-energy density batteries based on lithium ions. Recent experiments suggest the surprising result that the lithiation of a-Si is a two-phase process. Here, we present first-principles calculations of the structure of a-Si at different lithiation levels. Through a detailed analysis of the short and medium-range properties of the amorphous network, using Voronoi–Delaunay methods and ring statistics, we show that a-Li x Si has a fundamentally different structure below and above a lithiation level corresponding to x ∼ 2.
In polycrystalline materials, grain boundaries are sites of enhanced atomic motion, but the complexity of the atomic structures within a grain boundary network makes it difficult to link the ...structure and atomic dynamics. Here, we use a machine learning technique to establish a connection between local structure and dynamics of these materials. Following previous work on bulk glassy materials, we define a purely structural quantity (softness) that captures the propensity of an atom to rearrange. This approach correctly identifies crystalline regions, stacking faults, and twin boundaries as having low likelihood of atomic rearrangements while finding a large variability within high-energy grain boundaries. As has been found in glasses, the probability that atoms of a given softness will rearrange is nearly Arrhenius. This indicates a well-defined energy barrier as well as a well-defined prefactor for the Arrhenius form for atoms of a given softness. The decrease in the prefactor for low-softness atoms indicates that variations in entropy exhibit a dominant influence on the atomic dynamics in grain boundaries.