It seems that ML models (based on TMD or GNN) are not performing that well on the interneurons in datasets.
@lidakanari suggested that this may be due to the fact that some interneuron classes rely on directionality, similar to what happens for pyramidal cells pc of layers L2 and L6.
Actions
[ ] As we are already doing for pc-L2 and pc-L6, extract proj TMD and Graph features for in datasets
[ ] Run training on those new features and compare results.
Context
It seems that ML models (based on TMD or GNN) are not performing that well on the interneurons
indatasets.@lidakanari suggested that this may be due to the fact that some interneuron classes rely on directionality, similar to what happens for pyramidal cells
pcof layersL2andL6.Actions
pc-L2andpc-L6, extractprojTMD and Graph features forindatasets