Table 1.

Notations for MVHA.

Notation	Description
S, s, $s_k$	multivariate physiological signals (G and L), one of $g^{\left(cg\right)}$ or $l^{\left(cl\right)}$, the k-th segment in s
G, $g^{\left(cg\right)}$, $g_k$	high-frequency waveforms, the cg-th channel in G, the k-th segment in g
L, $l^{\left(cl\right)}$, $l_k$	numerical waveforms, the cl-th channel in L, the k-th segment in l
P∈ $R^{(U\times J)}$, $p^{\left(j\right)}$ ∈ $R^U$	the convolutional features, the j-th column in P
O∈ $R^{(U\times M)}$, o∈$R^U$, $o{f}_k$	output of the CNN layer, the sum of $p^{\left(j\right)}$, output of the fluctuant level attention
$\alpha$, ${\alpha }_k$, $\beta$, ${\beta }^{\left(k\right)}$	weights of the fluctuant level attention, the k-th value in $\alpha$, weights of the trend level attention, the k-th value in $\beta$
H$\in R^{(J\times M)}$, $h_k$	output of the Bi-LSTM layer, the k-th column in H
Z$\in R^{(J\times CH)}$, z	combination of H, the sum of $h_k$
X$\in R^{(I\times CH)}$, x	output of the fully connected layer, the k-th column of X
$A_{fl}$, $A_{tr}$	feature weights of the fluctuant level attention, feature weights of the trend level attention
d, $t{r}_{\left(kt\right)}$, $\rho$($s_k$)	output of the trend level attention, difference between $s_k$ and $s_{k-1}$, max, mean or min of $s_k$
$y_i$	prediction result of the i-th segment