Welcome!

Introduction to numerical algorithms and analytics as they apply to accuracy, efficiency, scalability, and stability within the fields of computing and engineering.  Current numerical computation trends, their limitations, advantages, and disadvantages will be discussed and modeled via problem-solving techniques.  Numerical computations will be done with Python.  

Course Page: https://umsystem.instructure.com/courses/262040/assignments/syllabus

Topics:

• Line search, Wolfe and Goldstein conditions
• Conjugate gradient methods
• Newton's methods
• Quasi-Newton
• Stochastic Gradient Descent
• Momentum
• Implementation using optim interface in PyTorch

Topics:

• Distributions, Fourier transform of different classes of functions and their properties.
• Linear operator, its convergence, and its spectra.
• Analyze the well-posedness of differential equation in a function spaces setup.