Data Science

The fundamental purpose of this course is to provide students with a robust interdisciplinary framework that goes beyond the application of algorithms, integrating theoretical physics and engineering with contemporary computational intelligence. This course seeks to transform the student into a scientific software architect and a probabilistic modeler capable of leading high-level research processes.

Module 1: Software Engineering for Reproducible Science

Students will be able to: (i) write clear and idiomatic Python for scientific work, (ii) structure research code as a maintainable project, (iii) use version control effectively, (iv) incorporate automated testing, (v) build reproducible environments, and (vi) profile and optimize bottlenecks while preserving correctness.

Module 2: Mathematical Probability for Data Analysis

Students will be able to: (i) model measurements with random variables and PDFs, (ii) estimate parameters and propagate uncertainty, (iii) use maximum likelihood as a unifying principle for estimation, (iv) incorporate systematic uncertainties, (v) perform hypothesis testing with appropriate statistics, and (vi) extend methods to multidimensional data.

Module 3: Machine Learning Techniques for Physicists and Engineers

Students will be able to: (i) formulate physics/engineering problems as ML tasks, (ii) build reproducible pipelines in Python, (iii) train and validate baseline and nonlinear models, (iv) quantify performance with appropriate metrics and uncertainty and calibration checks, and (v) interpret and subject models to stress tests under distributional and systematic shifts.

Faculty Community – Data Science