Skip to main content

Table 3 Average and standard deviations of BiLSTM-RSSI reconstruction method trained on GT1 after introducing random noise of various levels. Average values are computed over 20 independent realisations

From: Temporal social network reconstruction using wireless proximity sensors: model selection and consequences

Flipped random fraction

0.05

0.1

0.15

0.2

Average accuracy

90.02%

89.98%

89.91%

89.84%

Standard deviation

9.14%

9.24%

9.53%

9.26%