Table 3 Configuration information of different CNN models. The convolutional layer parameters are denoted as Conv < Layer i (1, 2, 3) >—< number of kernels > .
From: A hybrid deep neural network for classification of schizophrenia using EEG Data
A < 2,0,0 > | B < 2,2,0 > | C < 2,2,1 > | D < 4,2,1 > |
|---|---|---|---|
Input (\(32\times 32\) 3-channel EEG data) | |||
Conv1–32 | |||
Conv1–32 | Conv1–32 | Conv1–32 | Conv1–32 |
Conv1–32 | Conv1–32 | Conv1–32 | Conv1–32 |
Conv1–32 | |||
Max-pooling1 | |||
Conv2–64 | Conv2–64 | Conv2–64 | |
Conv2–64 | Conv2–64 | Conv2–64 | |
Max-pooling2 | |||
Conv3–128 | Conv3–128 | ||
Max-pooling3 | |||
FC-512 | |||
