Siddiqua Mazhar

Dr. Siddiqua Mazhar serves as Assistant Professor of Mathematics in the Division of Physical and Computational Sciences at the University of Pittsburgh. Her research pioneers mathematical frameworks for understanding deep learning through signal processing principles.

Her educational background includes a Ph.D. in Mathematics (2017) from Newcastle University specializing in Computational Group Theory. Prior to her current role, she worked as a Data Engineer at the University of Arizona and held adjunct teaching positions at Mid Michigan Community College and Southern New Hampshire University.

Dr. Mazhar's current research develops novel mathematical theories explaining neural network functionality using transform-domain sparse regularization, Radon transforms, and approximation theory. Her work critically examines weight decay regularization, skip connections, low-rank weight matrices, and sparsity mechanisms in high-dimensional problem solving.

She possesses extensive technical expertise in ELT processes, Python, SQL (MySQL, AWS Redshift), NoSQL (MongoDB), Power BI, Docker, Jira, Git, and AWS S3, enabling robust data pipeline development and machine learning implementation.

Dr. Mazhar teaches foundational mathematics courses including Calculus I-III, Linear Algebra, Abstract Algebra, Statistics and Probability with Python, Differential Equations, Group Theory, and Semigroup Theory.