The school will aim to familiarize participating students with state-of-the-art machine learning tools and frameworks, with a special focus on deep learning, to address challenges in computational mathematics and engineering. The focus will be on the theory and analysis of these methods, their practical implementation, and their effectiveness (and challenges) when used for solving real-world problems. The event will feature talks by leading experts in the field and provide opportunities for students to engage with them, share perspectives, and exchange ideas. Throughout the week, students will work on mentored, small-scale SciML projects and present their findings in a short presentation at the end of the school.

The Summer School will offer hands-on pre-training sessions over Zoom during the week prior to the main program, from July 29 to August 1, 2025. These sessions are designed to provide participants with a crash course in deep learning techniques and Python-based network design. Topics will include the fundamentals of neural network architectures, Physics-Informed Neural Networks (PINNs), operator networks, and generative AI models. Each day will feature a virtual lecture followed by a tutorial session to explore the code and concepts in more detail. These sessions will be particularly helpful in preparing participants for the projects during the main week of the Summer School. The full schedule, along with links to required materials and resources, will be available on the following webpage: https://sites.google.com/umd.edu/bmrc-sciml25-pretraining/home

The Zoom link for the sessions will be shared via email with the participants (including those selected for the primary week of the summer school from August 4 - August 8, 2025).