March 8, 2023
In Spring 2023, I will be teaching a three-credit course on the “Neural Network Methods for Signals in Engineering and Physical Sciences.” We will provide a practical introduction to Neural Networks and their application in the analysis of signal data common in engineering and physical sciences.
In the course, the fundamentals of artificial neural Networks will be surveyed. Along with fundamentals computational skills for working with and training neural networks will be built. In particular, instructions and assignments will introduce students with deep learning platforms: such as PyTorch, cloud GPU computing, and optimization algorithms: such as Stochastic Gradient Descent, Adam Dropout Initialization, etc. Students will build skills for working with two main types of deep neural networks: Convolutional Networks (CNNs) and Sequence Models (RNNs) which provide generic computational solutions for tasks concerning spatial or temporal (sequential) datasets, respectively.
In the second half of the course, students will work together in small groups and work toward projects with datasets across different disciplines composed by members of NSF HDR Institute A3D3: Accelerated Artificial Intelligence Algorithms for Data-Driven Discovery, such as High-Energy Physics, Multi-Messenger Astronomy, Gravitational Waves, Neuroscience, Audio and more.
The course is relevant to the Cross Cutting: Synthesis of Data Science Minor program. As a “project-based” course, students will have a hands-on experience with applications and limitations of neural network methods for signal analysis in engineering and applied sciences.
To enroll in this course, please use the course number PHYS 417.