Skip to main content

Advances, Systems and Applications

Table 4 Improved three-layer network model parameter settings

From: Low-cost and high-performance abnormal trajectory detection based on the GRU model with deep spatiotemporal sequence analysis in cloud computing

Training parameters

Parameter input

The number of hidden neurons in the first layer of network

10

The number of hidden neurons in the second layer network

10

The number of hidden neurons in the fully connected layer

24

Learning Rate

0.0006

Activation Function

sigmod/tanh

Iterations

100