Utilizing FLANN library for Multi-Label Classification

Question

I want to utilize the FLANN library for Mutli-Label Classification. I know that FLANN library is for computation of nearest neighbors, but I'm not sure how to utilize it for classification purposes. Is there some way to plug it in the Scikit-Learn or perhaps some other library.

Answer 1

FLANN is written in c++. However, the authors provide python bindings, such as this:

from pyflann import *
from numpy import *
from numpy.random import *
dataset = rand(10000, 128)
testset = rand(1000, 128)
flann = FLANN()
result,dists = flann.nn(dataset,testset,5,algorithm="kmeans",
branching=32, iterations=7, checks=16);

which I got from their manual.

---

That should get you started on how to combine it with scikit learn. The question on how to use NN for classification is too broad for an answer here, good luck!