|
52 | 52 | labels2 = labels1[n] |
53 | 53 |
|
54 | 54 | # Splitting the dataset into the Training set and Test set |
55 | | -X_train, X_valid, y_train, y_valid = train_test_split(data2, labels2, test_size=0.2, random_state=0) |
| 55 | +X_train, X_valid, y_train,y_valid = train_test_split(data2, |
| 56 | + labels2, test_size=0.2, random_state=0) |
56 | 57 | X_trainF = X_train.astype('float32') |
57 | | -X_validF = X_valid.astype('float32') |
| 58 | +X_validF = X_valid.astype('float32') |
58 | 59 | y_trainF = to_categorical(y_train) |
59 | 60 | y_validF = to_categorical(y_valid) |
60 | 61 |
|
61 | 62 | classifier = Sequential() |
62 | 63 | # CNN layers |
63 | | -classifier.add(Conv2D(32, kernel_size=(3, 3), input_shape=(36, 36, 3), activation='relu')) |
| 64 | +classifier.add(Conv2D(32, kernel_size=(3, 3), |
| 65 | + input_shape=(36, 36, 3), activation='relu')) |
64 | 66 | classifier.add(MaxPooling2D(pool_size=(2, 2))) |
65 | 67 | classifier.add(BatchNormalization(axis=-1)) |
66 | 68 | classifier.add(Dropout(0.5)) # Dropout prevents overfitting |
67 | | -classifier.add(Conv2D(32, kernel_size=(3, 3), input_shape=(36, 36, 3), activation='relu')) |
| 69 | +classifier.add(Conv2D(32, kernel_size=(3, 3), |
| 70 | + input_shape=(36, 36, 3), activation='relu')) |
68 | 71 | classifier.add(MaxPooling2D(pool_size=(2, 2))) |
69 | 72 | classifier.add(BatchNormalization(axis=-1)) |
70 | 73 | classifier.add(Dropout(0.5)) |
|
73 | 76 | classifier.add(BatchNormalization(axis=-1)) |
74 | 77 | classifier.add(Dropout(0.5)) |
75 | 78 | classifier.add(Dense(units=2, activation='softmax')) |
76 | | -classifier.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy']) |
77 | | -history = classifier.fit(X_trainF, y_trainF, batch_size=120, epochs=15, verbose=1, validation_data=(X_validF, y_validF)) |
| 79 | +classifier.compile(optimizer='adam', |
| 80 | + loss='categorical_crossentropy', metrics=['accuracy']) |
| 81 | +history = classifier.fit(X_trainF, y_trainF, |
| 82 | + batch_size=120, epochs=15, |
| 83 | + verbose=1, validation_data=(X_validF, y_validF)) |
78 | 84 | classifier.summary() |
79 | 85 |
|
80 | 86 | y_pred = classifier.predict(X_validF) |
|
0 commit comments