Creating a Training Set Pattern

What is an RFM Training Set?

The training set for RFM consists of two components: 

  1. a training set pattern (region of interest [ROI])
  2. at least 50 presence and 50 absence species validations 

Creating a training set allows you to add a Region of Interest (ROI) which is used by the model to create a training set pattern (ROI of the species' call). This article covers #1 (how to create a training set pattern). 

Note: Creating a training set applies to the Random Forest Model only. If you are interested in running Pattern Matching over your data, please reference Creating a Template instead. 

We recommend validating species in your recordings first, then once you find an ideal species call, add that as an training set pattern/ROI to your training set. 

How to Create a Training Set Pattern

In order to create a training set for a species, that species must already be added to your project. If you haven’t done this yet, reference this article on how to Add a Species

1. In the project menu click Audio analyses, then Random Forest Models. Select the ‘Training Sets’ tab and click the + icon. 

2. Assign a name and select the species + song type you want to create a training set for. Click Create. 

3. Open the Visualizer (project menu→Explore→Visualizer) and find a good example of your species’ call (strong signal with no overlapping sounds). We recommend using the most common call of the species (e.g. territorial song) to create the call template.

On the left Visualizer menu, click ‘Training Sets’.

4. Draw a bounding box around the species call. Click Add Data. 

Several ROIs can be added in the training set but we recommend using only one ROI per species. If using several ROIs, please make sure that the selected ROIs belong to the same species and call type and have similar acoustic properties (e.g. shape, bandwidth, duration). 

Next, you’ll need to validate the presence and absence of the focal species in a subset of recordings. Learn how to do it in our Validating Species help page.