def main():
hidden_unit = 64
batch_size = 32
learning_rate = 0.001
epochs = 50
with open('../Din/dataset/dataset.pkl', 'rb') as f:
train_set = np.array(pickle.load(f))
test_set = pickle.load(f)
cate_list = pickle.load(f)
user_count, item_count, cate_count, max_sl = pickle.load(f)
train_user, train_item, train_hist, train_sl, train_y = input_data(
train_set, max_sl)
# Tensorboard
current_time = datetime.datetime.now().strftime("%Y%m%d-%H%M%S")
log_dir = 'logs/' + current_time
tensorboard = tf.keras.callbacks.TensorBoard(log_dir=log_dir,
histogram_freq=1,
write_graph=True,
write_grads=False,
write_images=True,
embeddings_freq=0,
embeddings_layer_names=None,
embeddings_metadata=None,
embeddings_data=None,
update_freq=500)
# model checkpoint
check_path = 'save/wide_deep_weights.epoch_{epoch:04d}.val_loss_{val_loss:.4f}.ckpt'
checkpoint = tf.keras.callbacks.ModelCheckpoint(check_path,
save_weights_only=True,
verbose=1,
period=1)
model = WideDeep(user_count, item_count, cate_count, cate_list,
hidden_unit)
model.summary()
optimizer = tf.keras.optimizers.Adam(learning_rate=learning_rate)
model.compile(loss=tf.keras.losses.binary_crossentropy,
optimizer=optimizer,
metrics=[tf.keras.metrics.AUC()])
model.fit([train_user, train_item, train_hist, train_sl],
train_y,
epochs=epochs,
batch_size=batch_size,
validation_split=0.1,
callbacks=[tensorboard, checkpoint])
check_path = './save/deepfm_weight.epoch_{epoch:4d}.val_loss_{val_loss:.4f}.ckpt'
checkpoint = tf.keras.callbacks.ModelCheckpoint(check_path,
save_weights_only=True,
verbose=1,
period=5)
# ------------ model evaluate ------------
METRICS = [
tf.keras.metrics.BinaryAccuracy(name='accuracy'),
tf.keras.metrics.Precision(name='precision'),
tf.keras.metrics.Recall(name='recall'),
tf.keras.metrics.AUC(name='auc'),
]
model.compile(loss=tf.keras.losses.binary_crossentropy,
optimizer=tf.keras.optimizers.Adam(learning_rate=learning_rate),
metrics=METRICS)
# ---------早停法 -----
early_stopping = tf.keras.callbacks.EarlyStopping(monitor='val_auc',
verbose=1,
patience=10,
mode='max',
restore_best_weights=True)
model.fit(
train_X,
train_y,
epochs=epochs,
# callbacks=[early_stopping, checkpoint],
batch_size=batch_size,
sample_num=sample_num,
test_size=test_size)
train_X, train_y = train
test_X, test_y = test
# ============================Build Model==========================
model = WideDeep(feature_columns,
hidden_units=hidden_units,
dnn_dropout=dnn_dropout)
model.summary()
# ============================model checkpoint======================
# check_path = '../save/wide_deep_weights.epoch_{epoch:04d}.val_loss_{val_loss:.4f}.ckpt'
# checkpoint = tf.keras.callbacks.ModelCheckpoint(check_path, save_weights_only=True,
# verbose=1, period=5)
# ============================Compile============================
model.compile(loss=binary_crossentropy,
optimizer=Adam(learning_rate=learning_rate),
metrics=[AUC()])
# ==============================Fit==============================
model.fit(
train_X,
train_y,
epochs=epochs,
callbacks=[
EarlyStopping(monitor='val_loss',
patience=1,
restore_best_weights=True)
], # checkpoint
batch_size=batch_size,
validation_split=0.1)
# ===========================Test==============================
print('test AUC: %f' % model.evaluate(test_X, test_y)[1])
import pickle
from sklearn.metrics import log_loss, roc_auc_score
import tensorflow as tf
from builtins import int
# from loss import auc
model = WideDeep()
def auc(y_true, y_pred):
return tf.py_func(roc_auc_score, (y_true, y_pred), tf.double)
#model.compile(optimizer='rmsprop', loss=losses.mse, metrics=["mse"],)
model.compile(optimizer="adam",
loss="binary_crossentropy",
metrics=[auc, 'accuracy'])
#
# wide_features=pd.read_csv('data/path_matrix.txt', sep=' ', header=None)
# deep_features=pd.read_csv('data/sns_dense.csv', sep=',', header=0)
#
# for index, row in wide_features.iterrows():
# if not np.any(row):
# row[:]=8
# print('normalize')
#
# maxinterval=deep_features['interval'].max()
# print(maxinterval)
# for index, row in deep_features.iterrows():
# if row['interval'] == 0:
# row['interval'] = maxinterval