User talk:Jrm228/Benchmark III: Results

From BIOL 300 Wiki
Jump to: navigation, search

Student Comments

You should be more clear about which Izhikevich Model that you were comparing to Hodgkin-Huxley Model. For example, In figure 5, were you replicating a regular spiking neuron or a fast spiking neuron? It will be more clear if you mention that in the caption of figure 5.
Since you have mentioned how three parameters affect Izhikevich Model, I would recommend you to also discuss how the same parameters affect Hodgkin-Huxley Model. --Jingyi Yang (talk) 23:48, 24 April 2017 (EDT)

A few things: try adjusting the replicated figure parameters to show the full behavior published by Izhikevich. Also I think the real differences between the model come from not just the differences in the shape on a single action potential (which Izh. admits is simply lost data) but the differences in spike frequency between the HH and simple models when the parameters and current inputs are the same (i.e. both are regular spiking). There, you can really tell that the simple model does lose more than simple AP detail and actually affects the behavioral pattern, which is what we are trying to prove (even though our hypothesis is trying to disprove it).

Otherwise, nice! Just needs a bit more beef. Also some parameter explanation may fit better in the model description section. --Jessica Nash (talk) 12:04, 25 April 2017 (EDT)

Instructor Comments

  • Benchmark submitted on time?
    • Yes
  • Rubric submitted on time?
    • Yes
  • Results described and compared to original paper?
    • Yes, except for the interactions between the neurons.
  • Have you addressed the original biological hypothesis?
    • To some extent.
  • Figures and legends to show results?
    • Yes
  • Discrepancies relative to original model? Yes
  • Uploaded Mathematica file?
    • Yes

On the whole, this is a good to very good first draft of the term paper Results section. The most serious omission is that you have not shown any network interactions between the model neurons, which was the focus of the Izhikevich reduction. Showing even a relatively small network of interactions would be very useful. The comparison to the Hodgkin Huxley model is OK, but would be stronger if you focused more quantitatively on features of the Izhikevich model that do or do not replicate what is shown in the Hodgkin Huxley model. For example, can the Izhikevich model capture the current versus firing frequency characteristics of the Hodgkin Huxley model in at least one of its versions? Showing a dynamical analysis of the Izhikevich model will be an important extension.

--Hillel Chiel (talk) 20:39, 23 April 2017 (EDT)