Short-term forecasting of rail transit passenger flow based on long short-term memory neural network

Electricity supply chain hybrid long‐term demand forecasting approach based on deep learning—a case study of basic metals industry

“…LSTM consist of five main components: memory block, memory cells, input gate, output gate, and forget gate. The whole computation can be defined by a series of equations as follows [39]: $$$$\begin{equation}{i_t} = \sigma \left( {{W^i}H + {b^i}} \right)\end{equation}$$$$ $$$$\begin{equation}{f_t} = \sigma \left( {{W^f}H + {b^f}} \right)\end{equation}$$$$ $$$$\begin{equation}{o_t} = \sigma \left( {{W^o}H + {b^o}} \right)\end{equation}$$$$ $$$$\begin{equation}{c_t} = \tanh \left( {{W^c}H + {b^c}} \right)\end{equation}$$$$ $$$$\begin{equation}{m_t} = {f_t} \odot \,\,{m_{t - 1}} + {i_t} \odot \,\,{c_t}\end{equation}$$$$ $$$$\begin{equation}{h_t} = \tanh \left( {{o_t} \odot {m_t}} \right)\end{equation}$$$$σis the sigmoid function, $${W^i}$$,$${W^f}$$,…”

An Insight of Deep Learning Based Demand Forecasting in Smart Grids

“…This type of RNN contains an input layer, a recurrent hidden layer, and an output layer, with a memory block structure as shown in Figure 3 [ 70 ].…”

“…The LSTM memory block can be described according to the following equations [ 70 ]: where x t is the model input at time t ; W i , W f , W c , W 0 , U i , U f , U c , U 0 , V 0 are weight matrices; b i , b f , b c , b 0 are bias vectors; i t , f t , 0 t are respectively the activations of the three gates at time t ; c t is the state of memory cell at time t ; h t is the output of the memory block at time t ; ⊙ represents the scalar product of two vectors; is the gate activation function; g(x) is the cell input activation function; h(x) is the cell output activation function.…”

Short-term Passenger Flow Prediction of railway epidemic Based on SARIMA - LSTM combined model

“…In recent years, nonlinear models represented by neural networks have developed rapidly and have been widely used in pattern recognition, sample classification, prediction and other fields. For example, S. Sha [8] proposed a subway passenger flow prediction method based on RNN, but with the With the continuous improvement of relevant theoretical knowledge, we found that RNN cannot handle the problem of long-distance dependence very well; in 2018, Yuan Liu and Yong Qin [9] applied the LSTM model to short-term passenger flow prediction to solve the problem of RNN gradient disappearance during training. question.…”