the.com/cholesky factorization

square-rooting a matrix so a whole system of equations gets cheaper to solve.

means a way to break a symmetric positive-definite matrix A into L times L-transpose, where L is lower triangular, so solving Ax=b becomes two easy triangular solves instead of one hard one.

from named after andre-louis cholesky, a french military engineer who developed the method for geodesic surveying and artillery calculations before world war one, and never published it himself; a colleague released it posthumously in 1924 after cholesky died in combat in 1918.

for instance

kalman filtersused to factor covariance matrices at every update step

monte carlo simulationgenerates correlated random variables in quant finance models

gaussian process regressionfactors the kernel matrix to invert it cheaply

structural engineering feasolves stiffness matrix systems in finite element analysis

the.com/
what’s happening now · the.com · generated