Abbr.	Description	Abbr.	Description
ITS	Intelligent transportation system	$f_l$	LSTM feature space
TTP	Travel time prediction	$f_g$	GRU feature space
ATIS	Advanced traveler information system	$d_l$	LSTM decision space
GPS	Global positioning system	$d_g$	GRU decision space
OD	Origin–destination	$y_h$	Final output
GRU	Gated recurrent unit	$i^t$	Input gate
XGB	Extreme gradient boosting	$f^t$	Forget gate
LightGBM	Light gradient boosting machine	$o^t$	Output gate
FCD	Floating car data	${\sigma }^s$	Sigmoid activation function
MLP	Multilayer perceptron	$W^i$	Input gate weight
CNN	Convolutional neural network	$W^f$	Forget gate weight
BiLSTM	Bidirectional long short-term memory	$W^o$	Output gate weight
BiGRU	Bidirectional gated recurrent unit	$b^i$	Input gate bias
SVR	Support vector regressor	$b^f$	Forget gate bias
RMSE	Root mean square error	$b^o$	Output gate bias
MAE	Mean absolute error	$h^{t-1}$	Hidden state of prior timestamp
$R2$	Coefficient of determination	$x^t$	Current input
PCA	Principal component analysis	$C^t$	Cell state
DSAE	Deep stacked autoencoder	$h^t$	Hidden output
LSTM	Long short-term memory	${\mu }^{t^{\prime }}$	Tanh activation function
OSRM	Open-source routing machine	$W^c$	Cell state weight
$u^t$	Update gate	$b^c$	Cell state bias
${h^{\prime }}^t$	Current memory content time (t)	$r^t$	Reset gate
$h^t$	Final memory content time (t)	$W^r$, $U^r$	Reset gate weight
$W^u$, $U^u$	Update gate weight	⊙	Element-wise multiplication
$h^{f_t}$	Hidden state variable (Forward)	$h^{b_t}$	Hidden state variable (Backward)
$o^t$	Output layer state variable	$W^{f_o}$	Hidden output weight (Forward)
$W^{b_i}$	Hidden input weight (Backward)	$W^{f_i}$	Hidden input weight (Forward)
$W^{b_h}$	Hidden weight (Backward)	$W^{f_h}$	Hidden weight (Forward)
$W^{b_o}$	Hidden output weight (Backward)	f	Hidden layer activation
g	Output layer activation	TT_i	Actual travel time
$\widehat{T{T}_i}$	Predicted travel time	TT_m	Mean travel time
MI	Mutual information	E	Entropy
F	Feature	T	Target