Validating Presence of Species
A training set for the Random Forest Model (RFM) consists of two components: 1) a training set pattern (also called a template) and 2) at least 50 presence and 50 absence validations. This article covers how to validate species in your recordings. Next you'll want to read how to create a training set which allows you to create a training set pattern for the model to use to learn from.
Tip: Random Forest Models run optimally on both Firefox and Chrome.
1. Click on the Species Presence Validation section of the Visualizer tab and then click on the species category (e.g. Amphibians, Birds, Mammals, etc.).
2. Click Play on the Spectrogram to listen and determine if the species is present or absent. Click on Annotation to indicate presence or absence of a species call. Repeat these steps for a recordings’ subset. Subsets comprising of at least 50 recordings with the species call, and 50 without the species call are recommended to train Random Forest Models.