Block 4 — Scalar Kalman Filter#
Builds the predict-update cycle in one dimension, frames the Kalman gain as a trust factor between prediction uncertainty and measurement uncertainty, and walks through innovation, gain, and covariance updates on a 1D train example.
What you’ll learn#
Explain how the Kalman filter performs recursive optimal fusion.
Identify the roles of prediction, measurement, and innovation.
Interpret the Kalman gain as an uncertainty-weighted trust factor.
Update both the state and its uncertainty at every step.
Run the full predict-update cycle on a one-dimensional example.