Skip to main content
. 2023 Jun 23;13:10221. doi: 10.1038/s41598-023-34795-4

Figure 4.

Figure 4

Comparison of GRU and LSTM architectures’ performance for HS3D dataset. These networks are performing suboptimally compared to their bidirectional counterparts, and they are also unable to learn any distinguishing features in some of the experiments. Columns from left to right are (a) Accuracy per epoch, (b) F1 Score per epoch, and (c) AUR-PR per epoch.