Syllabus

 

Date

Lecture topics

Lab topics

Reading

 

Jan 3

(MS)

 

  • Course overview

  • Properties of time series

  • Data transformations

  • Time series decomposition

 

 

CM09: Chap 1

 

Jan 5

(MS)

 

  • Covariance & correlation

  • Autocorrelation & Partial autocorrelation

  • Cross correlation

  • White noise

 

  • Matrices in R & matrix algebra

  • Linear regression in matrix form

  • Basic matrix and ts functions in R

  • ACF & PACF for model identification

 

CM09:

Chaps 2 & 4

 

Jan 10

(MS)

 

  • Random walks

  • Autoregressive (AR) models

  • Moving average (MA) models

 

 

CM09:

Chaps 4 & 6

 

Jan 12

(EW)

 

  • Model estimation

  • Maximum likelihood

  • Bayesian estimation for this course

 

  • Simulating & fitting ARMA(p,q) models

  • Bayesian estimation via JAGS and STAN

 

 

Jan 17

(EH)

 

  • Univariate state-space models

  • Diagnostics for state-space models

   

 

Jan 19

(EH)

 

  • Introduction to multivariate state-space models

 

  • Fitting univariate & simple multivariate state-space models (MARSS)

 

HWS14:

Chaps 7 & 8

 

Jan 24

(EH)

 

  • Including covariates (predictors) in models

  • Seasonal effects

  • Missing covariates

  • Colinearity

 

 

HWS14Chap 12

 

Jan 26

(EW)

 

  • Multi-model inference and selection:

    • model selection metrics besides AIC

    • Cross-validation

  • Forecast performance metrics

 

  • Fitting multivariate state-space models with and without covariates

  • Model selection

  • Model diagnostics

 

 

Jan 31

(MS)

 

 

  • Univariate & multivariate dynamic linear models (DLMs)

 

 

Lamon et al. (1998)

HWS14Chap 15

 

Feb 2

(MS)

 

 

  • Applications of dynamic linear models (DLMs)

 

  • Fitting univariate & multivariate DLMs

 

 

Feb 7

(EH)

 

  • Forecasting with exponential smoothing models

  • More forecast assessment

 

 

Hyndman & Athanasopoulos: Chap 7.

 

 

Feb 9

(MS)

 

  • Overview of dynamic factor analysis (DFA)

 

  • Fitting DFA models with and without covariates

 

Zuur et al. (2003)

HWS14: Chap 9

Ohlberger et al. (2016)

 

 

Feb 14

(EW)

 

 

  • Overview of Bayesian estimation

  • DLMs, DFA, MARSS, SS models

   

 

Feb 16

(EW)

 

  • Time series models with non-Gaussian errors

  • Non-normal response variables

 

  • Fitting models with non-Gaussian errors

  • Fitting models with excessive zeros

 

 

Feb 21

(EW)

 

 

  • Time series models with spatial autocorrelation

   

 

Feb 23

(EH)

 

B matrix estimation I

  • Intro to Gompertz models, aka AR(1) and ARX(1)

  • Estimating species/population interactions from time series data

 

Meeting with students

  • 3:00 Eric: Lucas, Eli: Melanie

  • 3:20 Eric: Brianne, Eli: Vincent

  • 3:40 Eli/Eric: Silvana

 

Ives et al (2003)

HWS14: Chap 13

 

Feb 28

(EH)

 

B matrix estimation II

  • Studying community dynamics and stability using MAR(1) models

 

 

Ives et al (2003)

Hampton et al (2013)

 

Mar 2

(EH/MS)

 

  • No lecture today (March 2nd).  Last lecture on perturbation analysis will be Tues March 7th.

  •  

 

  • 2:30 Surakshya, Melanie, Lukas

  • 3:00 Brianne, Mike, Silvana

  • 3:30 Vincent

 

 

Mar 7

 

  • Lecture: Perturbation analysis

   

 

Mar 9

 

  • Student presentations

 

  • Celebration!

 

 

Announcements

No announcements

Send questions about this workspace to Eli Holmes.