Research Article
Recognition of Emotions Using Multichannel EEG Data and DBN-GC-Based Ensemble Deep Learning Framework
Algorithm 1
Pseudocodes for training the RBM with glia chain.
| Input: a training sample x1, learning rate ε, glia effect vector g | | Output: weight matrix W, bias vector of the visible layer b, bias vector of the hidden layer c | | Start training | | Initialize v(0): v(0)= x1 (initialize state value of the visible layer v) | | for j = 1 : m (for all hidden units) | | (hj(0) = 1 | v(0)) = σ ( + cj) | | End for | | Extract a sample h(0) according to h(0)∼ (h(0) | v(0)) | | for i = 1 : (for all visible units) | | ( = 1 | h(0)) = σ (ΣjWijhj(0) + bi) | | End for | | Extract a sample v(1) according to v(1)∼ (v(1) | h(0)) | | for j = 1 : m (for all hidden units, calculate the output value without glia effect) | | hj(1)∗ = ΣiWij(1)+ cj | | End for | | update the glia effect vector g | | for j = 1 : m (for all hidden units, calculate the output value with glia effect) | | (1) = σ ((1)∗ + ) | | End for | | Update parameters | | W ⟵ W + ε ( (h(0) = 1 | v(0))v(0)T − p(h(1) = 1 | v(1))v(1)T) | | b ⟵ b + ε (v(0) − v(1)) | | c ⟵ c + ε ( (h(0) = 1 | v(0)) − (h(1) = 1 | v(1))) |
|