hi there,

can anyone advance some python code for non linear regression analysis (time dependent) using loihi nengoDL lib ?

thk U

hi there,

can anyone advance some python code for non linear regression analysis (time dependent) using loihi nengoDL lib ?

thk U

I’m not very familiar with the problem you are referring to (“non linear regression analysis”) so it’s hard to make recommendations here. We have NengoDL examples here, as well as additional Nengo examples here. Perhaps you can find something useful on these pages?

Hi Xchoo.

I’ve went through those examples.

Maybe if i elaborate a bit more…

I have a set of input parameters in a 2D matrix and I’m looking to output a curve with one output. The outputs are obtained from experimental data. They were already modelled using a MMF equation. However, those equations one cannot tell the origin of regression constants obtained.

Experimental data is time dependent.

I’m looking to spike neural networks, in particular using neuromorphic ANN to model all experimental data at once (i already know to be possible to model ) .

Can i do it using nengoDL ?

I’m not familiar with the problem you are describing, but I can say that if you are able to train your ANN in TensorFlow, then it should be possible to train a spiking neuron version in NengoDL.

I spoke to other Nengo developers and unfortunately, no one has worked with your type of regression analysis before.