performance - How to speed up Python code for running on a powerful machine? -

performance - How to speed up Python code for running on a powerful machine? -

i've completed writing multiclass classification algorithm uses boosted classifiers. 1 of main calculations consists of weighted to the lowest degree squares regression. main libraries i've used include:

statsmodels (for regression) numpy (pretty much everywhere) scikit-image (for extracting pig features of images)

i've developed algorithm in python, using anaconda's spyder.

i need utilize algorithm start training classification models. i'll passing approximately 7000-10000 images algorithm, each 50x100, in grayness scale.

now i've been told powerful machine available in order speed training process. , asked me "am using gpu?" , few other questions.

to honest have no experience in cuda/gpu, etc. i've ever heard of them. didn't develop code such thing in mind. in fact had (ignorant) impression machine automatically run code faster mediocre one, without having it. (apart writing regular code efficiently in terms of loops, o(n), etc).

is still possible code speeded virtue of beingness on high performance computer? or need modify create utilize of parallel-processing machine?

i afraid can not speed programme running on powerful computer. had issue while back. first used python (very slow), moved c(slow) , had utilize other tricks , techniques. illustration possible apply dimensionality reduction speed things while having reasonable accurate result, or mentioned using multi processing techniques.

since dealing image processing problem, lot of matrix operations , gpu sure great help. there nice , active cuda wrappers in python can use, not knowing much cuda. tried theano, pycuda , scikit-cuda (there should more since then).

python performance numpy cuda gpu