[Biopython-dev] Neural Networks

Brad Chapman chapmanb at arches.uga.edu
Tue Mar 5 10:15:39 EST 2002


> OK, I will read your docs and try to understand what you did and I will
> organize my code to fit it.

Sounds great -- let me know if things don't make sense, or you need help
or whatever.

> From my initial tests it looks like the NN is very slow with the huge amount
> of data generated by sparse coding. I am not sure python is suitable for
> this but we can always give it a try. Worst case scenario is to use it for
> translating biological data to fit the input and translating the output back
> to biological data. The actual calculation can then be done by a C based
> engine.

Yup, python is definitely not speedy for neural networks. The optimal
solution would be to determine if the current framework I have makes
sense and, if so, code up the slow parts in C. Alternatively you could
hook into an existing NN engine, but I don't know much about what is
available. I worked in python so I could learn Neural Network stuff,
which is why I didn't choose to go this route. 

Let me know how things work out for you.
Brad



More information about the Biopython-dev mailing list