Instructor: Mengyang Gu Email: mengyang.gu, "at", pstat.ucsb.edu
Lecture time: Mon/Wed 2:00pm--3:15 pm, Location: ILP 3209
Office hours: Wednesday 12:50pm - 1:50pm, Location: South Hall 5511
________________________________________________________________________________________________________________________
Week Slides Topics Code/Extra slides/HW
01/06, 01/08 L1 L2 Introduction, key ideas in UQ, C1.1 C1.2 C1.3 C1.4 C1.5 R1 HW1
probabilistic models, Bayesian inference
01/13, 01/15 L3 Gaussian process surrogate, parameter estimation, C2.1 C2.2 C2.3 R2 R3 HW2
predictive distribution, RobustGaSP package
01/20 01/22 L4 Vectorized Gaussian processes, C3.1
Parallel partial Gaussian processes
01/27, 01/29 L5 Space-filling design, active learning strategies C4.1 C4.2
02/03, 02/05 L6 Covariance function, reproducing kernel Hilbert space,
deep neural network, comparison
02/10, 02/12. L7, Dynamic linear model and Kalman filter C6.1 C6.2
02/17, 02/19 L8 Implementation, dlm package C7.1 C7.2
02/24, 02/26 L9 State space representation of Gaussian processes,
PCA, dynamic mode decomposition, generalization
03/03, 03/05 L10 L11 Generalized probabilistic principal component analysis,
Fast data inversion of spatial and spatio-temporal data
03/10, 03/12 L12 L13 Imperfect model calibration R4 C10.1
03/20 Final report submission