Table 2 The AUROC score on different RNN cell unit.

From: An Automatic Assessment System for Alzheimer’s Disease Based on Speech Using Feature Sequence Generator and Recurrent Neural Network

 

Unidirectional

Bidirectional

GRU

LSTM

Simple

biGRU

biLSTM

biSimple

thr = 1

0.932 ± 0.02

0.922 ± 0.01

0.682 ± 0.08

0.951 ± 0.02

0.948 ± 0.02

0.871 ± 0.06

thr = 3

0.941 ± 0.03

0.916 ± 0.02

0.758 ± 0.06

0.969 ± 0.01

0.942 ± 0.01

0.888 ± 0.03

thr = 5

0.936 ± 0.02

0.922 ± 0.03

0.760 ± 0.03

0.948 ± 0.01

0.937 ± 0.02

0.885 ± 0.02