What are Linear time MMD and Quadratic time MMD in Shogun
In this recipe, we shall learn what are Linear time MMD and Quadratic time MMD in Shogun with the help of simple examples.
Recipe Objective: What are Linear time MMD and Quadratic time MMD in Shogun?
This recipe explains what Linear time MMD and Quadratic time MMD in Shogun are.
For more related projects-
Project 1
Project 2
Learn About the Application of ARCH and GARCH models in Real-World
Linear Time MMD
The linear time MMD implements a nonparametric statistical hypothesis test to reject the null hypothesis that to distributions p and q, only observed via n samples are the same.
# Linear Time MMD
MMD = LinearTimeMMD()
w = 1000
k = GaussianKernel(10, 1)
MMD.set_kernel(k)
MMD.set_p(p)
MMD.set_q(q)
MMD.set_num_samples_p(w)
MMD.set_num_samples_q(w)
a = 0.05
Quadratic Time MMD
The quadratic time MMD implements a nonparametric statistical hypothesis test to reject the null hypothesis that distributions p and q, only observed via n and m samples, respectively, are the same.
# Quadratic Time MMD
p = RealFeatures(f_p)
q = RealFeatures(f_q)
MMD = QuadraticTimeMMD(p, q)
k = GaussianKernel(10, 1)
MMD.set_kernel(k)
MMD.set_statistic_type(ST_BIASED_FULL)
stats = MMD.compute_statistic()
MMD.set_null_approximation_method(NAM_PERMUTATION)
MMD.set_num_null_samples(200)
val = MMD.compute_p_value(stats)
What Users are saying..
Abhinav Agarwal
I come from Northwestern University, which is ranked 9th in the US. Although the high-quality academics at school taught me all the basics I needed, obtaining practical experience was a challenge.... Read More