kNN.jl
kNN.jl copied to clipboard
johnmyleswhite
The k-nearest neighbors algorithm in Julia
Hi All, I just installed kNN using Pkg.add("kNN") and got the following error: ERROR: UndefVarError: kernelregression not defined If I use Pkg.checkout("kNN"), i get an error when trying to load...
The version of kNN pointed to by 0.0.0 in METADATA.jl/kNN is still `using Stats` rather than `using StatsBase`, which causes `using kNN` to fail with: ``` ERROR: Stats not found...
What wrong in model = knn(X, y, metric = Euclidean()) ERROR: function knn does not accept keyword arguments ? _ _ _ _(_)_ | A fresh approach to technical computing...
Classifier can accept data structures form [NearestNeighbors](https://github.com/johnmyleswhite/NearestNeighbors.jl) package.
FLANN (http://www.cs.ubc.ca/research/flann/) is one of the most widely used library for approximate nearest neighbor search. It is fast & reliable, available in Linux distro & Homebrew, and has a C...
There's no reason this package should depend on DataFrames.
It's very useful to apply k-nearest neighbors to things like text using purely text-based distance metrics.
Right now, `kernelregression` only works for functions that map R to R. It should also work for functions that map R^k to R.
Right now, `predict` produces the most probable class as its output. We should also expose a function like `classprobs` that produces the per-class probabilities for each class.