Total variational distance between two empirical distributions (KDE)
1. Computing empirical distribution For simplicity, we will assume that the data lies in two dimensional space with x1 and x2 axes. ndata1 = 100; ndata2 = 200; data1 = repmat([1 2], ndata1, 1) + randn(ndata1, 2); data2 = repmat([2 1], ndata2, 1) + randn(ndata2, 2); figure(1); hold on; plot(data1(:, 1), data1(:, 2), 'ro'); plot(data2(:, 1), data2(:, 2), 'bx'); axis equal; 2. Kernel Density Estima..
더보기
