Standardization is optional when performing PCA.
PCA can be used to visualize high-dimensional data in 2D or 3D.
LDA is a supervised learning technique.
The first principal component explains the most variance in the data.
PCA can be applied directly to categorical data without preprocessing.
PCA is affected by the order of the input features.
The principal components are orthogonal to each other.
sbox = (0xc, 0x4, 0x6, 0x2, 0xa, 0x5, 0xb, 0x9, 0xe, 0x8, 0xd, 0x7, 0x0, 0x3, 0xf, 0x1)
data = (0x0b, 0x1f, 0x04, 0x0d, 0x0a, 0x02, 0x17, 0xae, 0xff, 0x03, 0x05, 0x91, 0x01, 0x67, 0x0d, 0x0f)
dst = _mm_shuffle_epi8 (sbox, data);
Суперскалярний процесор містить:
GPU містять більше ядер ніж CPU