Improving the accuracy of polar motion prediction using a hybrid least squares and long short-term memory model

Earth, Planets and Space

Table 4 Experiments parameters

LSTM model parameters	Values
Total number of hidden layers	2
Number of LSTM cells per hidden layer	50
Time steps	365
Training iterations	1000
Dropout	0.1
Learn Rate Drop Factor	0.2
Loss function	MSE
Optimization algorithm	Adam
Smoothing method	SG smoothing filter
Gradient threshold	1