The goal of this project is to take a set of continuous time series data that are observed on the same time points, and to model the relationships among them in order to develop a prediction model. There is one time series of key interest in the collection which is a key outcome for prediction.
In this project we will use Vector Autoregression (VAR) to fit models with lagged time series as predictors.
- Learn the basics of time series modelling, including prediction and the evaluation of predictions
- Select, assemble and clean an appropriate data set
- Display the data, and carry out an exploratory analysis of each time series properties (in the ARIMA framework)
- Choose an outcome and search through a set of VAR models to find the best fit model
- Make a prediction using the model
- Make suitable displays of the model and the prediction.