Skip to main content

Table 4 Loss of Regression w/o L2 normalization after 200 epochs

From: EKF-based self-attitude estimation with DNN learning landscape information

  Dataset#
#1 #2 #3 #4
Loss [\({\text{m}}^2 {\text{/s}}^4\)] 0.6588 0.4916 1.0951 0.8284