Skip to main content

Table 2 Key symbols used in Forward BLSTM

From: Optimizing protein sequence classification: integrating deep learning models with Bayesian optimization for enhanced biological analysis

Symbol

Description

\(\:{\varvec{f}}_{\varvec{t}}^{\left(\varvec{f}\right)}\)

Forget the gate at the time step for the forward LSTM.

σ

Sigmoid activation function

\(\:{\varvec{W}}_{\varvec{f}}^{\left(\varvec{f}\right)}\)

Weight matrix for the forget gate

\(\:{\varvec{h}}_{\varvec{t}-1}^{\left(\varvec{f}\right)}\)

Hidden state at the previous time step t − 1for the forward LSTM

\(\:{\varvec{x}}_{\varvec{t}}^{\:}\)

Input at time step t

\(\:{\varvec{b}}_{\varvec{f}}^{\left(\varvec{f}\right)}\)

Bias term for the forget gate

\(\:{\varvec{I}\varvec{n}\varvec{p}\varvec{u}\varvec{t}}_{\varvec{t}}^{\left(\varvec{f}\right)}\)

Input gate at time step t for the forward LSTM

\(\:{\varvec{W}}_{\varvec{i}\varvec{n}\varvec{p}\varvec{u}\varvec{t}}^{\left(\varvec{f}\right)}\)

Weight matrix for the input gate

\(\:{\varvec{b}}_{\varvec{i}\varvec{n}\varvec{p}\varvec{u}\varvec{t}}^{\left(\varvec{f}\right)}\)

Bias term for the input gate

\(\:{\varvec{b}}_{\varvec{o}\varvec{u}\varvec{p}\varvec{u}\varvec{t}}^{\left(\varvec{f}\right)}\)

Output gate at time step t for the forward LSTM

\(\:{\varvec{W}}_{\varvec{o}\varvec{u}\varvec{t}\varvec{p}\varvec{u}\varvec{t}}^{\left(\varvec{f}\right)}\)

Weight matrix for the output gate

\(\:{\varvec{b}}_{\varvec{o}\varvec{u}\varvec{t}\varvec{p}\varvec{u}\varvec{t}}^{\left(\varvec{f}\right)}\)

Bias term for the output gate

\(\:{\check {\varvec{C}}}_{\varvec{t}}^{\left(\varvec{f}\right)}\)

Candidate cell state at time step t for the forward LSTM

tanh

Hyperbolic tangent activation function

\(\:{\varvec{W}}_{\varvec{C}}^{\left(\varvec{f}\right)}\)

Weight matrix for the candidate cell state

\(\:{\varvec{b}}_{\varvec{C}}^{\left(\varvec{f}\right)}\)

Bias term for the candidate cell state

\(\sigma\)

Cell state at time step t for the forward LSTM

\(\:{\varvec{C}}_{\varvec{t}-1}^{\left(\varvec{f}\right)}\)

Cell state at the previous time step t − 1t-1t − 1 for the forward LSTM

\(\:{\varvec{h}}_{\varvec{t}}^{\left(\varvec{f}\right)}\)

Hidden state at time step t for the forward LSTM