model.py

import scipy.io
import keras
from keras.models import Sequential
from keras.layers import Dense, Dropout, Flatten
from keras.layers import Conv2D, MaxPooling2D

# (get the data from files)
# read train data into python from .mat file
mat = scipy.io.loadmat('classifier_data.mat')
x_train = mat['TrainData_2d']
y_train = keras.utils.to_categorical(mat['Train_Lables'], num_classes=10)
# read validation data into python from .mat file
x_val = mat['ValData_2d']
y_val = keras.utils.to_categorical(mat['Val_Lables'], num_classes=10)
# read test data into python from a .mat file
x_test = mat['TestData_2d']
y_test = keras.utils.to_categorical(mat['Test_Lables'], num_classes=10)

num_classes = 10


# first conv layer config
conv1_batch = 30
conv1_window = (5, 5)
# second conv layer config
conv2_batch = 15
conv2_window = (3, 3)

model = Sequential()
model.add(Conv2D(conv1_batch, conv1_window, data_format='channels_first', input_shape=(1, 20, 146), activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Conv2D(conv2_batch, conv2_window, activation='relu'))
model.add(MaxPooling2D(pool_size=(2, 2)))
model.add(Dropout(0.2))

model.add(Flatten())
model.add(Dense(128, activation='relu'))
model.add(Dense(50, activation='relu'))
model.add(Dense(num_classes, activation='softmax'))

# Compile model [adam optimizer]
model.compile(loss='categorical_crossentropy', optimizer="adam", metrics=['accuracy'])

model.fit(x_train, y_train, validation_data=(x_val, y_val), batch_size=32, epochs=10, verbose=1, shuffle=True)
performance = model.evaluate(x_test, y_test, batch_size=32, verbose=0)