Machine Learning & AI

Keras Basics

Building neural networks with Keras API.

Import Keras

Import from TensorFlow
from tensorflow import keras
from tensorflow.keras import layers

Or standalone Keras
import keras
from keras import layers

Sequential Model

Create sequential model
model = keras.Sequential([
  layers.Dense(64, activation="relu", input_shape=(784,)),
  layers.Dropout(0.3),
  layers.Dense(10, activation="softmax")
])

Add layers incrementally
model = keras.Sequential()
model.add(layers.Dense(128, activation="relu"))
model.add(layers.Dense(10))

Common Layers

Dense layer
layers.Dense(64, activation="relu")

Dropout layer
layers.Dropout(0.5)

Convolutional layer
layers.Conv2D(32, (3, 3), activation="relu")

Max pooling
layers.MaxPooling2D((2, 2))

Flatten layer
layers.Flatten()

LSTM layer
layers.LSTM(128, return_sequences=True)

Compile & Train

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

Train model
history = model.fit(
  x_train, y_train,
  epochs=10,
  batch_size=32,
  validation_split=0.2
)

Evaluate
loss, acc = model.evaluate(x_test, y_test)