Exemplo n.º 1
0
    return model


text_class_model = build_model(113)

earlyStopping = keras.callbacks.EarlyStopping(monitor='loss',
                                              patience=10,
                                              verbose=1,
                                              mode='auto')

text_class_model.fit(train_data,
                     train_labels,
                     epochs=50,
                     batch_size=5,
                     callbacks=[earlyStopping])

ml_utils.save_model(text_class_model, 'text_class_model.h5')

test_predicted_res = text_class_model.predict(test_data, batch_size=1)

print(
    '\n****************Classification result for text classification************************'
)
ml_utils.display_result(
    test_labels_raw, test_predicted_res.argmax(axis=1),
    'text classification')  # Print the classification result

# for result in test_predicted_res:
#    ml_utils.display_confidence(result)
Exemplo n.º 2
0
    model.add(Convolution2D(nb_filters, nb_conv, nb_conv))
    model.add(Activation('relu'))

    model.add(MaxPooling2D(pool_size=(nb_pool, nb_pool)))
    model.add(Dropout(0.25))

    model.add(Flatten())
    model.add(Dense(128))
    model.add(Activation('relu'))
    model.add(Dropout(0.5))
    model.add(Dense(nb_classes))
    model.add(Activation('softmax'))

    model.compile(loss='categorical_crossentropy', optimizer='adadelta')

    model.fit(train,
              train_label,
              batch_size=batch_size,
              nb_epoch=nb_epoch,
              verbose=1,
              validation_data=(validation, validation_label))

    logger.info("model training complete")

    score = model.evaluate(validation, validation_label, verbose=0)

    logger.info("validation score: %f" % (score))

    save_model(model)
    logger.info("model saved")
Exemplo n.º 3
0
  model = Sequential()
  model.add(Convolution2D(nb_filters, nb_conv, nb_conv, border_mode = 'valid', input_shape = (1, img_rows, img_cols)))
  model.add(Activation('relu'))

  model.add(Convolution2D(nb_filters, nb_conv, nb_conv))
  model.add(Activation('relu'))

  model.add(MaxPooling2D(pool_size=(nb_pool, nb_pool)))
  model.add(Dropout(0.25))

  model.add(Flatten())
  model.add(Dense(128))
  model.add(Activation('relu'))
  model.add(Dropout(0.5))
  model.add(Dense(nb_classes))
  model.add(Activation('softmax'))

  model.compile(loss = 'categorical_crossentropy', optimizer = 'adadelta')

  model.fit(train, train_label, batch_size = batch_size, nb_epoch = nb_epoch, verbose = 1, validation_data = (validation, validation_label))

  logger.info("model training complete")

  score = model.evaluate(validation, validation_label, verbose = 0)

  logger.info("validation score: %f" % (score))

  save_model(model)
  logger.info("model saved")

Exemplo n.º 4
0
        earlyStopping = keras.callbacks.EarlyStopping(monitor='val_loss',
                                                      patience=10,
                                                      verbose=1,
                                                      mode='auto')

        model_rack = build_model(5)
        history_rack = model_rack.fit(train_rack_data,
                                      train_rack_labels,
                                      epochs=10,
                                      validation_data=(test_rack_data,
                                                       test_rack_labels),
                                      batch_size=500,
                                      verbose=2,
                                      callbacks=[earlyStopping])
        ml_utils.save_model(model_rack, 'model_rack.h5')

        model_container = build_model(25)
        start_time = time.time()
        history_container = model_container.fit(
            train_container_data,
            train_container_labels,
            epochs=50,
            validation_data=(test_container_data, test_container_labels),
            batch_size=1000,
            verbose=2,
            callbacks=[earlyStopping])
        elapsed_time = time.time() - start_time
        print('Deep Learning Training time: {}'.format(elapsed_time))
        ml_utils.save_model(model_container, 'model_container.h5')
                                  epochs=20,
                                  validation_data=(test_left_data,
                                                   test_left_labels),
                                  batch_size=500,
                                  verbose=2)

    model_right = dl_clf.build_model(8, input_shape)
    history_right = model_right.fit(train_right_data,
                                    train_right_labels,
                                    epochs=20,
                                    validation_data=(test_right_data,
                                                     test_right_labels),
                                    batch_size=500,
                                    verbose=2)

    ml_utils.save_model(model_bottom, 'model_bottom.h5',
                        config_save_load_dir_path)
    ml_utils.save_model(model_left, 'model_left.h5', config_save_load_dir_path)
    ml_utils.save_model(model_right, 'model_right.h5',
                        config_save_load_dir_path)

test_predicted_bottom_res = model_bottom.predict(test_bottom_data,
                                                 batch_size=1)
print(
    '\n****************Classification result for Bottom************************'
)
ml_utils.display_result(test_bottom_labels_raw,
                        test_predicted_bottom_res.argmax(axis=1),
                        'Bottom')  # Print the classification result

# for result in test_predicted_bottom_res:
#    ml_utils.display_confidence(result)
Exemplo n.º 6
0
    # earlyStopping = keras.callbacks.EarlyStopping(monitor='val_loss', patience=10, verbose=1, mode='auto')

    model_bottom = dl_clf.build_model(5, 6)
    history_bottom = model_bottom.fit(train_bottom_data, train_bottom_labels, epochs=10,
                                      validation_data=(test_bottom_data, test_bottom_labels), batch_size=500, verbose=2)

    model_left = dl_clf.build_model(7, 6)
    history_left = model_left.fit(train_left_data, train_left_labels, epochs=10,
                                  validation_data=(test_left_data, test_left_labels), batch_size=500, verbose=2)

    model_right = dl_clf.build_model(8, 6)
    history_right = model_right.fit(train_right_data, train_right_labels, epochs=10,
                                    validation_data=(test_right_data, test_right_labels), batch_size=500, verbose=2)

    ml_utils.save_model(model_bottom, 'model_bottom.h5')
    ml_utils.save_model(model_left, 'model_left.h5')
    ml_utils.save_model(model_right, 'model_right.h5')

test_predicted_bottom_res = model_bottom.predict(test_bottom_data, batch_size=1)
print('\n****************Classification result for Bottom************************')
ml_utils.display_result(test_bottom_labels_raw, test_predicted_bottom_res.argmax(axis=1),
                        'Bottom')  # Print the classification result

# for result in test_predicted_bottom_res:
#    ml_utils.display_confidence(result)

test_predicted_left_res = model_left.predict(test_left_data, batch_size=1)
print('\n****************Classification result for Left************************')
ml_utils.display_result(test_left_labels_raw, test_predicted_left_res.argmax(axis=1),
                        'Left')  # Print the classification result