pandinosaurus / pykernels

Python library for working with kernel methods in machine learning

• authors: Wojciech Marian Czarnecki and Katarzyna Janocha
• version: 0.0.4
• dependencies: numpy, scipy, scikit-learn

General description

Python library for working with kernel methods in machine learning. Provided code is easy to use set of implementations of various kernel functions ranging from typical linear, polynomial or rbf ones through wawelet, fourier transformations, kernels for binary sequences and even kernels for labeled graphs.

Sample usage

from sklearn.svm import SVC
from sklearn.metrics import accuracy_score
import numpy as np

from pykernels.basic import RBF

X = np.array([[1,1], [0,0], [1,0], [0,1]])
y = np.array([1, 1, 0, 0])

print 'Testing XOR'

for clf, name in [(SVC(kernel=RBF(), C=1000), 'pykernel'), (SVC(kernel='rbf', C=1000), 'sklearn')]:
    clf.fit(X, y)
    print name
    print clf
    print 'Predictions:', clf.predict(X)
    print 'Accuracy:', accuracy_score(clf.predict(X), y)
    print

implemented Kernels

• Vector kernels for R^d

  • Linear
  • Polynomial
  • RBF
  • Cosine similarity
  • Exponential
  • Laplacian
  • Rational quadratic
  • Inverse multiquadratic
  • Cauchy
  • T-Student
  • ANOVA
  • Additive Chi^2
  • Chi^2
  • MinMax
  • Min/Histogram intersection
  • Generalized histogram intersection
  • Spline
  • Sorensen
  • Tanimoto
  • Wavelet
  • Fourier
  • Log (CPD)
  • Power (CPD)

• Graph kernels

  • Labeled

    • Shortest paths

  • Unlabeled

    • Shortest paths
    • 3,4-Graphlets
    • Random walk