I am working on an edge impulse project. ‘‘motion recognition’’. I’ve created the model with, edge impulse studio. Now I’m trying a new aproch using the python SDK. I would like to use the dataset that I collected from the old project.
I downloaded the x_training.npy, y_training.npy, x_test.npy, y_test.npy docs from edge impulse studio.
But i got problems to understand the targets data (y_trainig, y_test) configuration.
In fact my model (classification model) has 21 features, and 4 classes (possible targets).
The inputs (x_train and x_test) config are fine : 21 columns one culomn for each feature. but why does the output has 4 columns?