User Tools

Site Tools


Translations of this page:

Back to BIS-Schools

Biodiversity data analysis with R






W08-1 Non-linear prediction

This worksheet revisits the regression and prediction topic but this time from a non-linear point of view. After completing this worksheet you should know how to use local regression models for fitting relationships within your data sets.

Things you need for this worksheet

  • R — the interpreter can be installed on any operation system. For Linux, you should use the r-cran packages supplied for your Linux distribution. If you use Ubuntu, this is one of many starting points. If you use windows, you could install R from the official CRAN web page.

  • R Studio — we recommend to use R Studio for (interactive) programming with R. You can download R Studio from the official web page.

  • your script and data from W02-1: Reading CSV files

Learning log assignments

:!: First things first: the following analysis is build on top of your script from W02-1. Please copy your script “W02-1.R”, rename the copy to “W08-1.R” and use it for the programming tasks of this worksheet.

:-\ Please visualize once again the relation between animal activity and coverage.

:-\ Perhaps there are models that fit better than the linear regression! Let's try a polynomial regression using the loess() function. Add the prediction of the loess model in our scatterplot using the lines() function which works almost identical to regLine().

:-\ Now let's check out how the loess model compares to the linear regression when it comes to predictions. Please compute a leave-one-out validation as in W06-1 but this time use the loess model. How do the error statistics compare to the linear prediction model?

en/learning/schools/s01/worksheets/ba-ws-08-1.txt · Last modified: 2015/09/22 16:22 (external edit)