Special Topics in Mathematical and Computational Sciences

Enrollment Requirements: Undergraduate Prerequisite: (CDSDS 122 or CASMA 581) and CDSDS 320 - Grade of C or higher required in all courses; Graduate Prerequisite: Graduate Standing Spring 2027 Topic - Machine Learning for Inverse Problems Question-driven Course Description: Inverse problems are inference problems where the aim is to infer an object or a quantity of interest from indirect and limited observations. Inverse problems arise in many data-driven problems in science and engineering. For instance, in medical imagery researchers are often interested in inferring the image of a tissue from noisy measurements from an imaging device such as an MRI machine. In public health, researchers are often interested in the dynamics of a disease from snapshot observations. In atmospheric science researchers may be interested in inferring atmospheric greenhouse gases from remote sensing data acquired by a satellite instrument. Inverse problems are characterized by a fundamental loss of information that prevents the exact recovery of the object of interest. Some amount of a-priori knowledge must be introduced to solve the problem, and the uncertainty in the proposed solution must be properly quantified. This class is a hand-on course that introduces the students to these topics.