def model_fn():
# build Keras Model
# use feature_column as keras input
input = {}
for f in FEATURE_NAME:
if f != TARGET:
input[f] = Input(shape=(1, ), name=f, dtype=DTYPE[f])
feature_columns = build_features()
feature_layer = tf.keras.layers.DenseFeatures(feature_columns)
dense_feature = feature_layer(input)
fm = FM_Layer(name='fm_layer', factor_dim=8)(dense_feature)
tf.summary.histogram('fm_output', fm)
output = Dense(1, activation='sigmoid', name='output')(fm)
model = Model(inputs=[i for i in input.values()], outputs=output)
optimizer = tf.keras.optimizers.Adam(learning_rate=0.01,
beta_1=0.9,
beta_2=0.999,
amsgrad=False)
model.compile(optimizer=optimizer,
loss='binary_crossentropy',
metrics=['binary_accuracy', 'AUC'])
print(model.summary())
return model
def model_fn(features, labels, mode, params):
"""
FM model
"""
feature_columns= build_features()
input = tf.feature_column.input_layer(features, feature_columns)
input_dim = input.get_shape().as_list()[-1]
with tf.variable_scope('linear'):
init = tf.random_normal( shape = (input_dim,1) )
w = tf.get_variable('w', dtype = tf.float32, initializer = init, validate_shape = False)
b = tf.get_variable('b', shape = [1], dtype= tf.float32)
linear_term = tf.add(tf.matmul(input,w), b)
add_layer_summary( linear_term.name, linear_term)
with tf.variable_scope('fm_interaction'):
init = tf.truncated_normal(shape = (input_dim, params['factor_dim']))
v = tf.get_variable('v', dtype = tf.float32, initializer = init, validate_shape = False)
sum_square = tf.pow(tf.matmul(input, v),2)
square_sum = tf.matmul(tf.pow(input,2), tf.pow(v,2))
interaction_term = 0.5 * tf.reduce_mean(sum_square - square_sum, axis=1, keep_dims= True)
add_layer_summary(interaction_term.name, interaction_term)
with tf.variable_scope('output'):
y = tf.math.add(interaction_term, linear_term)
add_layer_summary(y.name, y)
return y