DCT Lexicon · Readiness and stability
Kalman filter (belief update)
The recursion that carries a linear twin's belief: predict through the law, correct by the reading weighted by the gain.
Vol. II, Glossary of Key Terms, p. 399, its first sentence
In depth
In the book
The recursion that carries a linear twin’s belief: predict through the law, correct by the reading weighted by the gain (Definition 15.2, display (15.1)). Exact for Gaussian noises, the best linear estimate otherwise, its variance just after each reading settling at a fixed point (Theorem 15.1).
Vol. II, Glossary of Key Terms, p. 399, verbatim
Formal statement
Vol. II, Definition 15.2, display (15.1), retypeset
Common misreading
A common mistake is to read each new reading as the truth and discard what the law and earlier readings say.
Vol. II, LOS 15.2, verbatim
How AXIOM uses it
Each monthly upload is a reading. The Data refresh tab shows what the twin knew about quarter-end cash before the reading (the wide curve) and after it (the narrow one).
Drafted for the author's approval, from the Data refresh tab
In the book: Vol. II, Ch. 15: Enterprise Digital Twins and Autonomous Enterprise Transformation
Related: Belief-State Control · Monitor alarm (CUSUM) · Digital twin execution and the AXIOM laboratory
V2.Def.15.2 · V2.Thm.15.1 · GD.14


