Submit your question or comment.
Thank you very much for providing us with this excellent software.
Recently I have been studying “A nonlinear autoregressive Volterra model of the Hodgkin-Huxley equations”, the nonlinear autogressive method attracts me a lot. However, I have a question:
When training the model, how do you choose y_threshod, does it comes from the original H-H data, or calculated through the Volterra model once a time.
I have tried to reproduce this work for a long time, but I am alayways failed. I hope you can help me cope with it and reply to my email, thank you very much!
Recently, I have studied the methods you proposed in “A nonlinear autoregressive Volterra model of the Hodgkin–Huxley equations”, and we have reproduced the results in this article. After that we attempt to apply this method to a Hodgkin- Huxley type model invented by Rubin and Terman in 2004 (High Frequency Stimulation of the Subthalamic Nucleus Eliminates Pathological Thalamic Rhythmicity in a Computational Model),and we regulate the parameters, such as “Q, L, alpha, threshold”, according to the steps introduced in the article. When using the input-output data to train the model, we can get relatively good results, but when predicting the output in response to a novel input, we can’t obtain the satisfying results all the time. So I want to seek some advance from you.
1) Whether this method is suitable for different spiking patterns, such as bursting?
2) How to regulate the parameters more effectively?
Thank you very much! I hope for your reply!