2.174: Advancing Mechanics and Materials via Machine Learning

Spring | Graduate | 12 Units | Prereq: Permission of Instrutor

Concepts in mechanics (solid mechanics: continuum, micro, meso, andmolecular mechanics; elasticity, plasticity, fracture and buckling) and machinelearning (stochastic optimization, neural networks, convolutional neural nets, adversarial neural nets, graph neural nets, recurrent neural networks andlong/short-term memory nets, attention models, variational/autoencoders) introduced and applied to mechanics problems. Covers numerical methods, data and image processing, dataset generation, curation and collection, andexperimental validation using additive manufacturing. Modules cover: foundations, fracture mechanics and size effects, molecular mechanics andapplications to biomaterials (proteins), forward and inverse problems, mechanics of architected materials, and time dependent mechanical phenomena. Students taking graduate version complete additional assignments.