Forecasting fisheries catch time series with R (Forenoon)

Topics

  • Time-varying regression
  • Box-Jenkins (ARMA) Models
  • Exponential smoothing
  • Modelling time series with seasonality
  • Forecast diagnostics and accuracy metrics

Report-writing and code documentation with R (Afternoon)

Topics

  • Basic workflow using RStudio, Git and GitHub
  • Intro to R Markdown
  • Creating simple websites from RStudio
  • Build an R package with RStudio
  • Creating simple websites from R packages on GitHub
  • Creating a book with R Markdown: Intro to Bookdown.
  • Creating and publishing RShiny applications

Catch Forecasting Lectures and Labs

Introduction

Time-Varying Regression

Lectures Labs
1 Introduction to time-varying regression 1 Fit TV regression models to catch data
2 Forecasts with a time-varying regression model 2 Create time-varying regression forecasts

ARMA Models

Lectures Labs
1 Introduction to ARMA Models 1 Intro to ARMA models and diagnostic plots
2 Stationarity 2 Test the Greek catch data for stationarity
3 Selecting Model Structure 3 Fit ARMA Models to the Greek catch data
4 Fitting ARMA Models 4 Create and test forecasts
5 Create and test forecasts

Exponential Smoothing Models

Lectures Labs
1 Introduction to Exponential Smoothing Models 1 Fit exponential smoothing models to data
2 Selecting Model Structure 2 Create forecasts with exponential smoothing models
3 Forecasting with exponential smoothing models 3 Testing models

Seasonality

Lectures Labs
1 Introduction to seasonality and approaches 1 Creating time-series objects with seasonality in R
2 Seasonal time-vaying regression models 2 Seasonal exponential smoothing models
3 Seasonal exponential smoothing models