jnez71 / kalmann Goto Github PK

in Readme.md you write An implementation of chapter two from this text. but the link is broken.. so I dont know which text you are referring to. Could you possibly give the full reference ? Thanks

Hello,I see that in the knn.py, the conditions to check the convergence is "if len(RMS) > dslew and abs(RMS[-1] - RMS[-1-dslew])/dslew < dtol:",however, you set the value of dtol is -1, the value of "abs(RMS[-1] - RMS[-1-dslew])/dslew" must great than -1, why did you set this vaule of -1,and I want to consult you why "abs(RMS[-1] - RMS[-1-dslew])/dslew" is the condition of convergence.
thank you and waiting for your respond.
yours sincerely, Yannik Yang.

Make it easier for users to import module and install without having to work inside of the repo. @jnez71

We should be able to allow training to terminate with some termination criteria, regardless of the number of epochs requested. @jnez71?

Hi,

I am working with this for quite some time to replicate onto my own thesis, but I am having some troubles understanding how the Jacobian is calculated.

Can someone share some insight? Or post where I can read/learn about this.
It seems the H Jacobian seems to be a combination of stacks?

Lost - any help would be appreciated