skip to main content

Title: Predicting compositional changes of organic–inorganic hybrid materials with Augmented CycleGAN
Despite its simplicity, the composition of a material can be used as input to machine learning models to predict a range of materials properties. However, many property optimization tasks require the generation of novel but realistic materials compositions. In this study, we describe a way to generate compositions of hybrid organic–inorganic crystals through adapting Augmented CycleGAN, a novel generative model that can learn many-to-many relations between two domains. Specifically, we investigate the problem of composition change upon amine swap: for a specific chemical system (set of elements) crystalized with amine A, how would the product chemical compositions change if it is crystalized with amine B? By training with limited data from Cambridge Structural Database, our model can generate realistic chemical compositions for hybrid crystalline materials. The Augmented CycleGAN model can also utilize abundant unpaired data (compositions of different chemical systems), a feature that traditional supervised methods lack. The generated compositions can be used for many tasks, for example, as input fed to a classifier that predicts structural dimensionality.  more » « less
Award ID(s):
2018427 1928882
Author(s) / Creator(s):
; ;
Date Published:
Journal Name:
Digital Discovery
Page Range / eLocation ID:
255 to 265
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    Oxidation states (OS) are the charges on atoms due to electrons gained or lost upon applying an ionic approximation to their bonds. As a fundamental property, OS has been widely used in charge‐neutrality verification, crystal structure determination, and reaction estimation. Currently, only heuristic rules exist for guessing the oxidation states of a given compound with many exceptions. Recent work has developed machine learning models based on heuristic structural features for predicting the oxidation states of metal ions. However, composition‐based oxidation state prediction still remains elusive so far, which has significant implications for the discovery of new materials for which the structures have not been determined. This work proposes a novel deep learning‐based BERT transformer language model BERTOS for predicting the oxidation states for all elements of inorganic compounds given only their chemical composition. This model achieves 96.82% accuracy for all‐element oxidation states prediction benchmarked on the cleaned ICSD dataset and achieves 97.61% accuracy for oxide materials. It is also demonstrated how it can be used to conduct large‐scale screening of hypothetical material compositions for materials discovery.

    more » « less
  2. Abstract

    Generative models are a sub-class of machine learning models that are capable of generating new samples with a target set of properties. In chemical and materials applications, these new samples might be drug targets, novel semiconductors, or catalysts constrained to exhibit an application-specific set of properties. Given their potential to yield high-value targets from otherwise intractable design spaces, generative models are currently under intense study with respect to how predictions can be improved through changes in model architecture and data representation. Here we explore the potential of multi-task transfer learning as a complementary approach to improving the validity and property specificity of molecules generated by such models. We have compared baseline generative models trained on a single property prediction task against models trained on additional ancillary prediction tasks and observe a generic positive impact on the validity and specificity of the multi-task models. In particular, we observe that the validity of generated structures is strongly affected by whether or not the models have chemical property data, as opposed to only syntactic structural data, supplied during learning. We demonstrate this effect in both interpolative and extrapolative scenarios (i.e., where the generative targets are poorly represented in training data) for models trained to generate high energy structures and models trained to generated structures with targeted bandgaps within certain ranges. In both instances, the inclusion of additional chemical property data improves the ability of models to generate valid, unique structures with increased property specificity. This approach requires only minor alterations to existing generative models, in many cases leveraging prediction frameworks already native to these models. Additionally, the transfer learning strategy is complementary to ongoing efforts to improve model architectures and data representation and can foreseeably be stacked on top of these developments.

    more » « less
  3. null (Ed.)
    Deep Neural Networks (or DNNs) must constantly cope with distribution changes in the input data when the task of interest or the data collection protocol changes. Retraining a network from scratch to combat this issue poses a significant cost. Meta-learning aims to deliver an adaptive model that is sensitive to these underlying distribution changes, but requires many tasks during the meta-training process. In this paper, we propose a tAsk-auGmented actIve meta-LEarning (AGILE) method to efficiently adapt DNNs to new tasks by using a small number of training examples. AGILE combines a meta-learning algorithm with a novel task augmentation technique which we use to generate an initial adaptive model. It then uses Bayesian dropout uncertainty estimates to actively select the most difficult samples when updating the model to a new task. This allows AGILE to learn with fewer tasks and a few informative samples, achieving high performance with a limited dataset. We perform our experiments using the brain cell classification task and compare the results to a plain meta-learning model trained from scratch. We show that the proposed task-augmented meta-learning framework can learn to classify new cell types after a single gradient step with a limited number of training samples. We show that active learning with Bayesian uncertainty can further improve the performance when the number of training samples is extremely small. Using only 1% of the training data and a single update step, we achieved 90% accuracy on the new cell type classification task, a 50% points improvement over a state-of-the-art meta-learning algorithm. 
    more » « less
  4. null (Ed.)
    Abstract Accurate theoretical predictions of desired properties of materials play an important role in materials research and development. Machine learning (ML) can accelerate the materials design by building a model from input data. For complex datasets, such as those of crystalline compounds, a vital issue is how to construct low-dimensional representations for input crystal structures with chemical insights. In this work, we introduce an algebraic topology-based method, called atom-specific persistent homology (ASPH), as a unique representation of crystal structures. The ASPH can capture both pairwise and many-body interactions and reveal the topology-property relationship of a group of atoms at various scales. Combined with composition-based attributes, ASPH-based ML model provides a highly accurate prediction of the formation energy calculated by density functional theory (DFT). After training with more than 30,000 different structure types and compositions, our model achieves a mean absolute error of 61 meV/atom in cross-validation, which outperforms previous work such as Voronoi tessellations and Coulomb matrix method using the same ML algorithm and datasets. Our results indicate that the proposed topology-based method provides a powerful computational tool for predicting materials properties compared to previous works. 
    more » « less
  5. The Deep Neural Network (DNN) model is known for its high accuracy in classification tasks due to its intrinsic ability to learn the underlying patterns existing in a set of data. Hence it has gained momentum in seizure detection research, as in many other fields. However, its high performance is at the expense of an extensive training time. This is not appropriate for a real-time application such as seizure detection in which a swift reaction is required to save the life of the patient. This paper presents a novel Kriging-Bootstrapped Deep Neural Network hierarchical model for early seizure detection in which Kriging is first used to generate a well-correlated intermediate data set from the original input. The correlated data is then fed into the DNN for the final training. Experiments were carried out using electroencephalogram (EEG) data from both normal and epileptic patients. Results show that, with the same architecture and data size, the cumulative training time of the Krigging-Bootstrapped DNN is about 75% lower than that of the ordinary DNN without a compromise in performance as the proposed hybrid model shows a slightly better accuracy than the baseline DNN model. 
    more » « less