search

cog-imperial / OMLT

Represent trained machine learning models as Pyomo optimization formulations
Other
257 stars 56 forks source link

MaxPool layers: add support for non-increasing activation functions on preceding convolutional layer #154

Closed jezsadler closed 2 weeks ago

jezsadler commented 2 weeks ago

Per the TODO note on line 273 of full_space.py:

# TODO: add support for non-increasing activation functions on preceding
# convolutional layer
rmisener commented 2 weeks ago

We don't support non-increasing activation functions and don't have a plan to at this point.