Outlier-robust Kalman Filtering through gener
-
+
The dotted blue line shows the KF posterior mean estimate and
the solid orange line shows the WoLF posterior mean estimate.
@@ -167,7 +167,7 @@
Outlier-robust Kalman Filtering through gener
-Tresholded Mahalanobis-based weighting function (TMD)
+Thresholded Mahalanobis-based weighting function (TMD)
$$
W_{t}(\bm y_{1:t}) =
\begin{cases}
@@ -210,7 +210,7 @@
Outlier-robust Kalman Filtering through gener
Remarks
The Kalman filter is not outlier-robust.
-
The IMQ and TMD are outlier-robust.
+
Filters with IMQ and TMD weighting function are outlier-robust.
@@ -249,7 +249,7 @@
Outlier-robust Kalman Filtering through gener
KF-B
-
\(O(I_{-\epsilon}\,p^3)\)
+
\(O(I\,p^3)\)
3
Wang2018
@@ -279,8 +279,9 @@
Outlier-robust Kalman Filtering through gener
- Below, \(I\) is the number of inner iterations, \(I_{-\epsilon}\) is the number of inner iterations to reach a threshold \(\epsilon\),
- \(p\) is the dimension of the state vector, and #HP is the number of hyperparameters.
+ Below, \(I\) is the number of inner iterations,
+ \(p\) is the dimension of the state vector,
+ and #HP is the number of hyperparameters.