Difference between 'multi output' vs 'raw' in keras 'flow_from_dataframe'

user_4397712 0
Viewed 561

I'm not sure about when to use raw vs multi output in the keras flow_from_dataframe class_mode parameter, as by the looks of it, they both provide a way to classifying data with multiple labels. Suppose say I have a dataframe with the image path as well as two columns/classes with labels for each given image, and I would like to create a model which classifies the images based on those classes, which class_mode would I use, and when would I use the other one?

Edit: attached an image of the dataframe I'm using

enter image description here

keras tensorflow python
1 Answers

Use class_mode="raw" when the label column you're using has the actual raw class values that you intend to use as the training label. For example, if you're doing a regression task or ordinal regression and you have floating point numbers or integers as your columns. In that case, you must make sure that the actual numerical values are what you want your final labels to be.

In your case, it looks like you have text labels of different classes, so you have a multi-class, multi-label classification problem, and therefore you must use class_mode="multi_output" to transform the y-values correctly with multiple labels.

user_37916400
Related