Table 5 Summary of parameters and their descriptions.
From: Segmentation-enhanced approach for emotion detection from EEG signals using the fuzzy C-mean and SVM
Parameter/symbol | Description |
|---|---|
xi | i-th data point or feature vector |
cj | Center of the j-th cluster (used in FCM) |
uij | Membership degree of xix_i in cluster cjc_j (FCM) |
m | Fuzziness parameter in FCM controlling degree of overlap (typicallyā>ā1) |
w, b | Weight vector and bias term in SVM defining the decision hyperplane |
Ī¾i | Slack variable to allow soft-margin classification (SVM) |
C | Regularization parameter in SVM controlling margin vs. error trade-off |
Ī³ | Parameter in kernel functions controlling spread or influence |
Filters | Number of learnable filters in CNN convolutional layers |
Kernel size | Size of convolutional kernel (e.g., (3,3)) in CNN |
Pool size | Size of pooling window used in MaxPooling layers |
Units | Number of memory cells or neurons in the LSTM layer |
Activation | Activation function used in LSTM or CNN (e.g., ReLU, softmax) |
Epochs | Number of full training passes over the dataset |
Batch size | Number of training samples processed in each iteration |
Optimizer | Algorithm used to update model weights (e.g., Adam) |
Learning rate | Controls the step size in weight updates during training |
N estimators | Number of decision trees used in Random Forest |
Max depth | Maximum depth of any individual decision tree in the forest |
