This course will cover model-based time series analysis with a particular focus on applications in ecology and climatology. All methods will be illustrated using the free, open-source software package R.
Time Series data are ubiquitous in the physical sciences, and models for their behaviour enable scientists to understand temporal dynamics and predict future values.
The Course
Participants will be taught a wide range of suitable time series models for both discrete and continuous time systems. The course takes a foundational Bayesian approach, which will enable participants to have a deeper understanding of the models being fitted, and to estimate all unknown quantities with uncertainty. Participants are encouraged to bring their own data sets for discussion with the course tutors.
Content
Day 1 Basic concepts:
- Class 1: Introduction; some example time series datasets; prediction vs explanation
- Class 2: An introduction to Bayesian Statistics.
- Class 3: The AR(1) model
- Practical: revision on using R to load data, create plots and fit statistical models
- Round table discussion: understanding the output from a Bayesian model
Day 2 Arima modelling
- Class 1: ARMA models for real data
- Class 2: ARIMA and sARIMA modelling
- Practical: An introduction to the Bayesian modelling language JAGS
- Round table discussion: understanding and running a
- JAGS model
Day 3 Continuous Time Series Modelling
- Class 1: Brownian Motion and its application to real data sets
- Class 2: An introduction to Stochastic Volatility Modelling
- Practical: Fitting continuous time models in JAGS
- Round table discussion: Issues of continuous vs discrete time
Day 4 Advanced Times Series Models
- Class 1: Multivariate models
- Class 2: Fractional differencing and models using differential equations
- Practical: Running advanced models in JAGS
- Round table discussion: Bring your own data set
Cost
Course only ?440.00 + VAT
All-inclusive ?570.00 + VAT
Contact
Please email any enquiries or visit the website for more information