Topics in time series

In the spring 2017 semester, I taught a graduate course on topics in time series analysis, with focus on data and under financial support from Data@Carolina. Check out the syllabus of the course and the following lectures, organized by the themes covered and some given by students or guests.

Univariate

Classical decomposition

Differencing and unit roots

State space and structural modeling

Spectral perspective

Long memory

Seasonal and periodic models

Multivariate

Vector autoregressions (VARs)

Variable selection and penalized estimation

Cointegration

Graphical models and structural VARs

Factor models

Classification and clustering

Spatio-temporal modeling

Nonstationary

Change points

Time-frequency/Time-scale perspective

Wavelets

Locally stationary time series

Anomaly/Outlier detection

Nonlinear

Threshold autoregressive models

Count time series

Hidden Markov models

Neural network models