This course will introduce modern computational methods for materials science, including artificial intelligence (AI), machine learning (ML) and density functional theory (DFT). Both AI/ML and DFT are becoming standard tools in chemistry, physics, and materials science. Deep learning specifically involves linking input data (features) with output data (labels) through a neural network. Neural networks are capable of approximating any function. A typical example is the relationship between a material's structure and its properties. Conversely, DFT is a computational method used to analyze the electronic structure of atoms, molecules, and materials. It is based on quantum mechanics and provides valuable insights into the properties and behavior of various materials. Both DFT and AI/ML have their own strengths and applications, and they can be combined. Depending on the engineering field and the specific problem, these methods can be relevant. Therefore, AI/ML and DFT are valuable tools for students who will become engineers and scientists. Weekly Topics: 1 Introduce Python and install the computing environment. 2 Math review: Tensors and shapes 3 Introduction to machine learning (ML) 4 ML concepts: Regression, model assessment, classification, and kernel learning 5 Introduction to deep learning 6 Graph neural networks (GNN) 7 Equivariant neural networks 8 Mid-semester examination 9 Introduction to material science 10 Application of ML in material science 11 Introduction to density functional density (DFT) and Quantum ESPRESSO (QE) 12 Practical DFT with QE: Basic parameters 13 Practical DFT with QE: Advanced topics 14 Practical DFT with QE: Input generator 15 Combine DFT and ML 16 Final examination Assessments: * Mid-semester examination (50%): 3 hours * Final examination (50%): 3 hours

1)Understanding the DFT and AI/ML concepts. 2)Can practice the DFT and AI/ML by using the open-source Quantum ESPRESSO, TensorFlow, and Pytorch. 3)Using the dataset from the Material Project. Apply DFT and AL/ML for practical applications in material science, such as screening solar cell materials.

Need to bring your own laptop to class

2 hours

Textbook or Designated Reading: (1) S. Sandfeld, Materials Data Science - Introduction to Data Mining, Machine Learning, and Data-Driven Predictions for Materials Science and Engineering, Springer, (2024) (2) I. Goodfellow, Y. Bengio and A. Courville, Deep Learning, MIT Press, 800 Pages, (2016) (3) N. T. Hung, A. R. T. Nugraha and R. Saito, Quantum ESPRESSO Course for Solid‑State Physics, Jenny Stanford Publishing, New York, 372 Pages, (2022). Note: Additional journal articles, reviews, and materials will be given out in classes. Other References: - MIT Introduction to Deep Learning | 6.S191 https://www.youtube.com/playlist?list=PLtBw6njQRU-rwp5__7C0oIVt26ZgjG9NI - Quantum ESPRESSO for Solid State Physics https://nguyen-group.github.io/courses/qe/