Uncertainties in DiffEqFlux.jl

question

#1

I watched the JuliaCon 2018 | PuMaS.jl: Pharmaceutical Modeling and Simulation Engine given by @ChrisRackauckas and saw the part about propagation of parameter uncertainties around 21:47.

A (possibly) quick question regarding this in DiffEqFlux.jl and neural DEs. Is it possible to get uncertainties from the neural DE part? Especially uncertainties around an ANN in issue How to mix “normal” ODEs and Neuronal ODEs?, when it is implemented.

Thanks!


#2

Interesting question. Haven’t tried it. Let’s give it a try and get it working. In principle we could. Open an issue.


#3

Feel free to ping me in case you need support on the Measurements.jl side :slightly_smiling_face: