Entropy-Based Estimation of Event-Related De/Synchronization in Motor Imagery Using Vector-Quantized Patterns

. 2020 Jun 24;22(6):703. doi: 10.3390/e22060703

Algorithm 1 Building of VQ stochastic patterns.

1:
procedureVector Quatization in X
2:
Input: ${\tilde{x}}_{n} (τ, M; q), q \in [1, Q]$
3:
Initialize the reduced set ${\bar{X}}_{n} (τ, M)$ , then ${\bar{x}}_{n} (τ, M; 1) = {\tilde{x}}_{n} (τ, M; 1)$
4:
for $q \in [2, Q]$ do
5:
Compute the distance between ${\tilde{x}}_{n} (τ, M; q)$ and ${\bar{X}}_{n} (τ, M)$ . $d ({\tilde{x}}_{n} (τ, M; q), {\bar{X}}_{n} (τ, M)) = | | {\tilde{x}}_{n} (τ, M; q) - {\bar{x}}_{n} (τ, M; q^{'}) {| |}_{2}^{2}, q^{'} \in [1, Q^{'}]$
6:
if $| | d ({\tilde{x}}_{n} (τ, M; q), {\bar{X}}_{n} (τ, M)) > ρ {| |}_{1} = Q^{'}$ then
7:
${\bar{X}}_{n} (τ, M) \leftarrow {\tilde{x}}_{n} (τ, M; q)$
8:
$Q^{'} = Q^{'} + 1$
9:
end if
10:
end for
11:
end procedure