Data Analytics for Energy Systems

Catalog description

Data Analytics for Energy Systems merges introductory statistics with coding through simulation-based inference. Class lectures are split between discussing the concepts and illustrating their application through coding examples. The course is highly data-centric, using mostly datasets pertaining to the energy industry or grand challenges related to energy and sustainability (although some data are generated through fun in-class exercises). The course starts by discussing different types of data and introducing students to basic coding skills to manipulate datasets and extract summary statistics. The course then moves into data visualization, discussing common types of graphical tools and the types of data for which they are appropriate. Simulation-based inference then begins around the third week with bootstrapping in the context of finding confidence intervals, which then moves into hypothesis testing through randomization distributions. The concepts from hypothesis testing carry over into chi-square tests, ANOVA, and regression, which also delves into prediction. Neural Networks and Random Forests are covered at a conceptual and applied level (not getting into the theory) towards the end of the course. The course concludes with some topics in basic probability that weren't covered earlier in the semester.