Calendar
Week 1
- Sep 26
- Lecture 1
- HW 0 out
- Course outline, examples of time series data
- Sep 28
- Lecture 2
- HW 0 due HW 1 out
- CQ1 out: stationarity, ACF
- Course logistics, Models for time series data, stationarity, autocorrelation, and basic theoretical constructs
- Reading: S&S 1.2 - 1.4
Week 2
- Oct 3
- Lecture 3
-
- Estimating ACF; Data wrangling – detrending and data transformations
- Reading: S&S 1.5, 2.2
- Oct 5
- Lecture 4
-
- CQ1 due
- Data wrangling cont’d – smoothing via nonparametric regression; Linear and AR(1) processes (videos – Yom Kippur)
- Reading: S&S 2.3, 3.1
Week 3
- Oct 10
- Lecture 5
-
- AR, MA, and ARMA models
- Reading: S&S 3.1-3.2
- Oct 12
- Lecture 6
- HW 1 due HW 2 out
- CQ2 out: data wrangling, ARMA models
- ARMA ACF and partial ACF; Best linear prediction, forecasting
- Reading: S&S 3.3-3.4
Week 4
- Oct 17
- Lecture 7
-
- Forecasting cont’d, forecasting metrics; Estimation via method of moments
- Reading: S&S 3.4-3.5
- Oct 19
- Lecture 8
-
- CQ2 due
- Estimation via maximum likelihood and least squares, estimation of ARMA models; ARIMA/SARIMA models
- Reading: S&S 3.5-3.6
Week 5
- Oct 24
- Lecture 9
- Project Proposal due
- ARIMA and SARIMA diagnostics and model selection; VAR models
- Reading: S&S 3.7, 3.9, 5.6
- Optional Reading: Lutkepohl 2.1
- Oct 26
- Lecture 10
- HW 2 due HW 3 out
- CQ3 out: forecasting, estimation of ARMA, ARIMA/SARIMA
- VARMA models; State space models
- Reading: S&S 6.1
- Optional Reading: Lutkepohl 11.1-11.3 (VARMA), 3.1-3.4 (estimation), 18.2-18.2 (SSMs)
Week 6
- Oct 31
- Lecture 11
-
- Filtering, smoothing, and other state space models (DLFM, HMMs)
- Reading: S&S 6.2
- Optional Reading: Lutkepohl 18.3
- Nov 2
- Lecture 12
-
- CQ3 due
- Estimation of state space models
- Reading: S&S 6.3-6.5
- Optional Reading: Lutkepohl 18.4; PML2 4.2, 8.1-8.4, 30.1-30.4, 31.1-31.2
Week 7
- Nov 7
- Lecture 13
-
- Deep learning intro
- Reading: PML2 16.1-16.3
- Optional Reading: Lutkepohl 2.1
- Nov 9
- Lecture 14
- HW 3 due HW 4 out
- CQ4 out: VARMA/SSMs, filtering/smoothing, HMMs, deep learning
Week 8
- Nov 14
- Lecture 11
- Project Midway due
- Andy Miller (Apple) guest lecture
- Nov 16
- Lecture 12
-
- CQ4 due
- David Sontag (MIT) guest lecture
Week 9
- Nov 28
- Lecture 17
-
- Granger causality and neural Granger causality
- Nov 30
- Lecture 18
- HW 4 due
- CQ5 out: more on RNNs/CNNs, Granger causality, GP basics
Week 10
- Dec 5
- Lecture 19
-
- Dec 7
- Poster Session
-
- CQ5 due