|
|
import os
|
|
|
import sys
|
|
|
sys.path.append(os.getcwd())
|
|
|
import pickle
|
|
|
import pandas as pd
|
|
|
from tqdm import tqdm
|
|
|
from sklearn.preprocessing import LabelEncoder, MinMaxScaler
|
|
|
import warnings
|
|
|
from deepctr.models import xDeepFM
|
|
|
from deepctr.feature_column import SparseFeat, DenseFeat, get_feature_names
|
|
|
from tensorflow.keras import backend as K
|
|
|
from tensorflow.keras.layers import *
|
|
|
from tensorflow.keras.models import *
|
|
|
from tensorflow.keras.callbacks import *
|
|
|
from tensorflow.keras.models import save_model
|
|
|
import tensorflow as tf
|
|
|
from tensorflow.compat.v1 import ConfigProto
|
|
|
from tensorflow.compat.v1 import InteractiveSession
|
|
|
|
|
|
from config import offline_mode, subject_features_save_path
|
|
|
from config import subject_train_user_item_feats
|
|
|
from config import subject_val_user_item_feats
|
|
|
from config import subject_test_user_item_feats
|
|
|
from config import subject_model_save_path
|
|
|
from config import subject_all_user_item_feats
|
|
|
from config import subject_rank_dense_fea
|
|
|
from config import subject_rank_sparse_fea
|
|
|
from config import subject_rank_feats_columns
|
|
|
from config import logger
|
|
|
from matching.subject.recall_comm import get_item_info_df
|
|
|
|
|
|
os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
|
|
|
os.environ["CUDA_VISIBLE_DEVICES"] = "1"
|
|
|
|
|
|
K.set_learning_phase(True)
|
|
|
if tf.__version__ >= '2.0.0':
|
|
|
tf.compat.v1.disable_eager_execution()
|
|
|
|
|
|
tqdm.pandas()
|
|
|
warnings.filterwarnings('ignore')
|
|
|
|
|
|
# 加这几行避免训练报错
|
|
|
config = ConfigProto()
|
|
|
config.gpu_options.allow_growth = True
|
|
|
session = InteractiveSession(config=config)
|
|
|
|
|
|
def get_xdeepfm_feats_columns(df, dense_fea, sparse_fea, emb_dim=32):
|
|
|
"""
|
|
|
数据准备函数:
|
|
|
df: 数据集
|
|
|
dense_fea: 数值型特征列
|
|
|
sparse_fea: 离散型特征列
|
|
|
"""
|
|
|
fixlen_feature_columns = [SparseFeat(feat, vocabulary_size=df[feat].max() + 1, embedding_dim=emb_dim)
|
|
|
for i, feat in enumerate(sparse_fea)] + [DenseFeat(feat, 1, )
|
|
|
for feat in dense_fea]
|
|
|
|
|
|
dnn_feature_columns = fixlen_feature_columns
|
|
|
linear_feature_columns = fixlen_feature_columns
|
|
|
|
|
|
feature_names = get_feature_names(linear_feature_columns + dnn_feature_columns)
|
|
|
|
|
|
x = {}
|
|
|
x = {name: df[name].values for name in feature_names}
|
|
|
|
|
|
return x, linear_feature_columns, dnn_feature_columns
|
|
|
|
|
|
if __name__ == '__main__':
|
|
|
|
|
|
logger.info('获取物品信息')
|
|
|
subject_info_df = get_item_info_df()
|
|
|
|
|
|
# 课程总数量
|
|
|
subjects_count = len(subject_info_df['subject_id'].unique().tolist())
|
|
|
|
|
|
logger.info('加载用户行为特征')
|
|
|
|
|
|
# 所有用户物品特征
|
|
|
all_user_item_feats_df = pd.read_csv(subject_all_user_item_feats, sep='\t', encoding='utf-8')
|
|
|
|
|
|
train_user_item_feats_df = pd.read_csv(subject_train_user_item_feats, sep='\t', encoding='utf-8')
|
|
|
train_user_item_feats_df['subject_id'] = train_user_item_feats_df['subject_id'].astype(int)
|
|
|
|
|
|
if offline_mode:
|
|
|
val_user_item_feats_df = pd.read_csv(subject_val_user_item_feats, sep='\t', encoding='utf-8')
|
|
|
val_user_item_feats_df['subject_id'] = val_user_item_feats_df['subject_id'].astype(int)
|
|
|
else:
|
|
|
val_user_item_feats_df = None
|
|
|
|
|
|
test_user_item_feats_df = pd.read_csv(subject_test_user_item_feats, sep='\t', encoding='utf-8')
|
|
|
test_user_item_feats_df['subject_id'] = test_user_item_feats_df['subject_id'].astype(int)
|
|
|
|
|
|
# 做特征的时候为了方便，给测试集也打上了一个无效的标签，这里直接删掉就行
|
|
|
del test_user_item_feats_df['label']
|
|
|
|
|
|
# 把特征分开
|
|
|
sparse_fea = subject_rank_sparse_fea
|
|
|
|
|
|
dense_fea = subject_rank_dense_fea
|
|
|
|
|
|
train_user_item_feats_df[dense_fea] = train_user_item_feats_df[dense_fea].fillna(0, )
|
|
|
if val_user_item_feats_df is not None:
|
|
|
val_user_item_feats_df[dense_fea] = val_user_item_feats_df[dense_fea].fillna(0, )
|
|
|
test_user_item_feats_df[dense_fea] = test_user_item_feats_df[dense_fea].fillna(0, )
|
|
|
|
|
|
# dense特征进行归一化
|
|
|
for feat in dense_fea:
|
|
|
min_max_scaler = MinMaxScaler()
|
|
|
min_max_scaler.fit((all_user_item_feats_df[[feat]]))
|
|
|
pickle.dump(min_max_scaler, open(subject_model_save_path + 'min_max_scaler_' + feat + '.model', 'wb'))
|
|
|
|
|
|
train_user_item_feats_df[feat] = min_max_scaler.transform(train_user_item_feats_df[[feat]])
|
|
|
if val_user_item_feats_df is not None:
|
|
|
val_user_item_feats_df[feat] = min_max_scaler.transform(val_user_item_feats_df[[feat]])
|
|
|
test_user_item_feats_df[feat] = min_max_scaler.transform(test_user_item_feats_df[[feat]])
|
|
|
|
|
|
# sparse特征进行LabelEncoder
|
|
|
for feat in sparse_fea:
|
|
|
label_encoder = LabelEncoder()
|
|
|
if feat == 'subject_id':
|
|
|
label_encoder.fit(subject_info_df[[feat]])
|
|
|
subject_id_lable_encoder = label_encoder
|
|
|
else:
|
|
|
label_encoder.fit(all_user_item_feats_df[[feat]])
|
|
|
if feat == 'user_id':
|
|
|
user_id_label_encoder = label_encoder
|
|
|
pickle.dump(label_encoder, open(subject_model_save_path + feat + '_label_encoder.model', 'wb'))
|
|
|
|
|
|
train_user_item_feats_df[feat] = label_encoder.transform(train_user_item_feats_df[[feat]])
|
|
|
if val_user_item_feats_df is not None:
|
|
|
val_user_item_feats_df[feat] = label_encoder.transform(val_user_item_feats_df[[feat]])
|
|
|
test_user_item_feats_df[feat] = label_encoder.transform(test_user_item_feats_df[[feat]])
|
|
|
|
|
|
# 准备训练数据
|
|
|
x_train, linear_feature_columns, dnn_feature_columns = get_xdeepfm_feats_columns(
|
|
|
train_user_item_feats_df, dense_fea, sparse_fea)
|
|
|
y_train = train_user_item_feats_df['label'].values
|
|
|
|
|
|
if val_user_item_feats_df is not None:
|
|
|
# 准备验证数据
|
|
|
x_val, linear_feature_columns, dnn_feature_columns = get_xdeepfm_feats_columns(
|
|
|
val_user_item_feats_df, dense_fea, sparse_fea)
|
|
|
y_val = val_user_item_feats_df['label'].values
|
|
|
|
|
|
dense_fea = [x for x in dense_fea if x != 'label']
|
|
|
x_test, linear_feature_columns, dnn_feature_columns = get_xdeepfm_feats_columns(
|
|
|
test_user_item_feats_df, dense_fea, sparse_fea)
|
|
|
|
|
|
model = xDeepFM(linear_feature_columns, dnn_feature_columns, task='binary')
|
|
|
model.summary()
|
|
|
|
|
|
model.compile(optimizer="Adam",
|
|
|
loss=tf.keras.losses.BinaryCrossentropy(from_logits=True),
|
|
|
metrics=['AUC']#评价指标：AUC,精确度，均方误差
|
|
|
)
|
|
|
# 模型训练
|
|
|
if val_user_item_feats_df is not None:
|
|
|
history = model.fit(x_train,
|
|
|
y_train,
|
|
|
verbose=1,
|
|
|
epochs=5,
|
|
|
validation_data=(x_val, y_val),
|
|
|
batch_size=256)
|
|
|
else:
|
|
|
# 也可以使用下面的语句用自己采样出来的验证集
|
|
|
history = model.fit(x_train,
|
|
|
y_train,
|
|
|
verbose=1,
|
|
|
epochs=5,
|
|
|
validation_split=0.3,
|
|
|
batch_size=256)
|
|
|
|
|
|
# 保存训练好的XDepFM模型
|
|
|
save_model(model, subject_model_save_path + 'xdeepfm_model.h5')
|
|
|
|
|
|
# 模型预测
|
|
|
test_user_item_feats_df['pred_score'] = model.predict(x_test, verbose=1, batch_size=256)
|
|
|
|
|
|
# 还原user_id和subject_id
|
|
|
test_user_item_feats_df['user_id'] = user_id_label_encoder.inverse_transform(test_user_item_feats_df[['user_id']])
|
|
|
test_user_item_feats_df['subject_id'] = subject_id_lable_encoder.inverse_transform(test_user_item_feats_df[['subject_id']])
|
|
|
|
|
|
item_info_df = get_item_info_df()
|
|
|
item_info_df = item_info_df[['subject_id', 'subject_name']]
|
|
|
test_user_item_feats_df = test_user_item_feats_df.merge(item_info_df, how='left', on='subject_id')
|
|
|
|
|
|
test_user_item_feats_df.sort_values(by=['user_id', 'pred_score'], ascending=[True, False], inplace=True)
|
|
|
test_user_item_feats_df['pred_rank'] = test_user_item_feats_df.groupby(['user_id'])['pred_score'].rank(ascending=False, method='first').astype(int)
|
|
|
|
|
|
for feat in subject_rank_feats_columns:
|
|
|
min_max_scaler = pickle.load(open(subject_model_save_path + 'min_max_scaler_' + feat + '.model', 'rb'))
|
|
|
|
|
|
train_user_item_feats_df[feat] = min_max_scaler.inverse_transform(train_user_item_feats_df[[feat]])
|
|
|
if val_user_item_feats_df is not None:
|
|
|
val_user_item_feats_df[feat] = min_max_scaler.inverse_transform(val_user_item_feats_df[[feat]])
|
|
|
test_user_item_feats_df[feat] = min_max_scaler.inverse_transform(test_user_item_feats_df[[feat]])
|
|
|
|
|
|
test_user_item_feats_df[['user_id', 'subject_id', 'subject_name'] + subject_rank_feats_columns + ['pred_score', 'pred_rank']].to_csv(
|
|
|
subject_features_save_path + 'xdeepfm_rank_score.csv', sep='\t', index=False, header=True) |
