Summarized from CUHK(SZ) MAT2041(2022Fall) Lecture 17.
Motivation
Any n n n linear independent n n n-vector can form a basis of a n n n-dim space. However, people are accustomed to orthogonal bases. In high school students learn only Cartesian coordinate system. If we can transform a random basis into an orthogonal (even orthonormal) basis, it’s more convenient in some situations.
Gram Schmidt Process: to change a random basis into an orthonormal basis.
Solution
Assume that the ordinary basis is { u 1 , u 2 , … , u n } \{u_1,u_2,\ldots,u_n\} { u1,u2,…,un}, we want { v 1 , v 2 , … , v n } ( ∀ i ≠ j , v i T v j = 0 ; ∀ i , ∣ ∣ v i ∣ ∣ = 1 ) \{v_1,v_2,\ldots,v_n\}(\forall i\neq j,\ v_i^Tv_j=0;\forall i, ||v_i||=1) { v1,v2,…,vn}(∀i=j, viTvj=0;∀i,∣∣vi∣∣=1).
There’s no limitation to the first vector, for convenience, we just pick v 1 = u 1 ∣ ∣ u 1 ∣ ∣ v_1=\frac{u_1}{||u_1||} v1=∣∣u1∣∣u1.
Then we try to modify u 2 u_2 u2 to meet the requirements.
u 2 u_2 u2 should be orthogonal to v 1 v_1 v1, and v 2 v_2 v2 could be expressed by only v 1 v_1 v1 and u 2 u_2 u2.
Thus we assume ( u 2 − α v 1 ) ⊥ v 1 (u_2-\alpha v_1)\bot v_1 (u2−αv1)⊥v1, then v 1 T u 2 = α v 1 T v 1 = α v_1^Tu_2=\alpha v_1^Tv_1=\alpha