Neural Networks with Keras Basics

What command is used to install TensorFlow?

• pip install tensorflow

• tensorflow install

• install tensorflow

• pip install tf

How do you import Keras in a Python script?

• from keras.models import Sequential

• import keras as keras

• from keras import Sequential

• import keras.models.Sequential

What function is used to create a sequential model in Keras?

• createSequentialModel()

• KerasSequential()

• ModelSequential()

• Sequential()

Which method is used to compile a Keras model?

• fit

• evaluate

• compile

• train

What is the purpose of the fit() method in Keras?

• To evaluate a model's performance.

• To save the model architecture.

• To preprocess the input data.

• To train a model on a dataset.

How can you load a dataset using TensorFlow?

• Use the `tf.load` function to import datasets.

• Manually read files without any preprocessing.

• Use the `tf.data` API to create and preprocess datasets.

• Load datasets using NumPy only.

What is the role of the validation_split parameter in model training?

• The validation_split parameter determines the learning rate of the model.

• The validation_split parameter is used to increase the batch size during training.

• The validation_split parameter specifies the number of epochs for training.

• The validation_split parameter helps in setting aside a portion of training data for validation during model training.

How do you preprocess image data in TensorFlow?

• Convert images to grayscale before normalization.

• Apply random transformations without resizing the images.

• Use tf.data.Dataset to load images directly from a URL.

• Use tf.io.read_file(), tf.image.decode_image(), tf.image.resize(), and normalize pixel values.

What is the purpose of the reshape() function in data preprocessing?

• To analyze data for trends.

• To visualize data in charts.

• To filter out irrelevant data.

• To change the shape or structure of data for analysis.

How can you save a trained Keras model?

• model.save_weights('my_model_weights.h5')

• model.save_model('my_model.pkl')

• model.save('my_model.h5')

• model.export('my_model.json')