|| Over the past decade, novel artificial intelligence (AI) methods have been developed as predictive tools for materials science applications. It is important, however, that the predicted structures can be physically manufactured or synthesized. This symposium will feature advanced AI and data analytics approaches for the design of materials through composition engineering and optimization of available or novel processing techniques. Methods that investigate the impact of elemental composition, processing, and synthesis conditions on the material structure/properties are of particular interest. Examples of AI methods capable of building and exploiting processing-structure-property relationships include variational autoencoders, reinforcement learning, Bayesian optimization, evolutionary algorithms, mixed integer non-linear optimization algorithms, and advanced design of experiments approaches.
- Automated methods to optimize the composition, synthesis and processing conditions of materials including alloys, ceramics, composites, battery materials, catalysts, membranes and photovoltaics
- Materials design approaches leveraging experiments and physics-informed simulations in concert with advanced optimization strategies
- Application of materials informatics approaches including machine learning, deep learning, genetic algorithms, uncertainty quantification via bootstrapping or gaussian processes, and Bayesian optimization towards materials process optimization