Developing Forecast Models from Time-Series Data in MATLAB (Part 1)

Abhaya Parthy, MathWorks

Are you looking to increase your data analysis capabilities? Do you need to perform complex analytics and automate cumbersome repetitive tasks such as batch processing? Do you need to make your programs accessible to others?

During this presentation, we demonstrate how you can use MATLAB to develop nonlinear predictive models from historical time-series measurements. As a working case study, a forecast model of short-term electricity loads for the Australian market using BOM and AEMO data is presented. This case study applies nonlinear tree bagging regression and neural network modelling techniques. At the end of the case study, the MATLAB forecast model is converted into a deployable plug-in for Microsoft Excel.

Product Focus

  • MATLAB

Recorded: 9 Oct 2012