Explores phase transitions in physics and computational problems, highlighting challenges faced by algorithms and the application of physics principles in understanding neural networks.
Explores optimal errors in high-dimensional models, comparing algorithms and shedding light on the interplay between model architecture and performance.
Explores generalization in machine learning, focusing on underfitting and overfitting trade-offs, teacher-student frameworks, and the impact of random features on model performance.
Explores the application of machine learning in molecular dynamics and materials, emphasizing the creation of meaningful features and the importance of generalizability.
Explores applying machine learning to atomic scale systems, emphasizing symmetry in feature mapping and the construction of rotationally invariant descriptors.