Materials & Nanoscience

Data Mining and Machine Learning Meets Experiment and First-Principles Simulation for Materials Discovery (#314)

This symposium will focus on data-driven and physics-based approaches to accelerate the discovery and development of new functional materials, motivated by the United States’ Materials Genome Initiative and similar funding initiatives world-wide. This symposium will bring together experimentalists and theorists working in a wide range of organic and inorganic materials, with applications in gas storage and separation, thermoelectrics, photovoltaics, and electronic materials. The symposium will have three points of emphasis: (i) Databases that connect synthesis, property measurement, and theoretical calculations for materials screening and discovery; (ii) Data-mining/machine-learning approaches used for finding patterns in these experimental and theoretical databases, and developing predictive models of relevant physical properties and engineering figures of merit from this data; (iii) Integration of atomistic simulation methods with experimental materials discovery into the discovery process of (i) and (ii). In addition to contributions from researchers already in this area, we want to use this symposium to build collaborations that will promote public database construction among materials chemists who have currently-underutilized new datasets.
Last update: Dec 28, 2015