File size: 3,435 Bytes
9697a93 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
import pandas as pd
from sklearn.metrics.pairwise import cosine_similarity
from sklearn.metrics import mean_squared_error
from xgboost import XGBRegressor
def train_model(data,user_id, test=None, eval = False):
# select only user data
train_user = data[data['userId']==user_id]
X_train = train_user.drop(columns=['userId','rating', 'Train', 'title'])
y_train = train_user['rating']
model = XGBRegressor()
model.fit(X_train,y_train)
if eval:
test_user = test[test['userId']== user_id]
X_test = test_user.drop(columns=['userId','rating', 'Train', 'title'])
y_test = test_user['rating']
y_pred = model.predict(X_test)
rmse = np.sqrt(mean_squared_error(y_test, y_pred))
print(f'RMSE: {rmse:.4f}')
# Model evaluation
# print("Predected rating:", y_pred)
# print("Actual rating:",y_test)
# print(X_test)
return model
def get_user_recommendation_XGBoost(all_moves,model, user_id, n=10):
# get all movies that the user has not seen
user_seen_movies = all_moves[all_moves['userId'] == user_id]['title']
user_unseen_movies = all_moves[~all_moves['title'].isin(user_seen_movies)]
# drop duplicates
user_unseen_movies = user_unseen_movies.drop_duplicates(subset=['title'])
# make predictions
user_unseen_movies['Pred_rating'] = model.predict(user_unseen_movies.drop(columns=['userId', 'rating', 'Train', 'title']))
# only return movies with more than 100 ratings
# get top 10 recommendations
recommendations = user_unseen_movies.sort_values(by='Pred_rating', ascending=False).head(n)['title']
return recommendations ,user_seen_movies
def get_user_recommendation(DataBase, Matrix,user_id,l=10):
user = Matrix[user_id]
user = user.sort_values(ascending=False)
# now we have a series of user similarities
# we only want to recommend movies that the user has not seen
# so we need to filter out movies that the user has seen
user_seen_movies = DataBase[DataBase['userId'] == user_id]['title']
# Now we loop through user and get top 10 recommendations
recommendations = []
print(len(user.index))
for U in user.index[1:10]:
# get all rated movies by user U
movies = DataBase[DataBase['userId'] == U]['title']
movies = movies[~movies.isin(user_seen_movies)]
# get all movies that U has rated 4 or higher
movies = movies[DataBase['rating'] >= 4]
# sort by rating
movies = movies.sort_values(ascending=False)
for movie in movies[:4]:
if movie not in recommendations:
recommendations.append(movie)
if len(recommendations) >= l:
break
return recommendations
if __name__ == '__main__':
import pickle
def load_similarity_matrix(path):
with open(path, 'rb') as f:
similarity_df = pickle.load(f)
return similarity_df
# Load the data
DataBaseCSV = r"D:\Study\ITI\Recommender Systems\Final\Movies-Recommender-System\Data\XGBoost_database.csv"
DataBase = pd.read_csv(DataBaseCSV)
# Load the similarity matrix
MatrixCSV = r"D:\Study\ITI\Recommender Systems\Final\Movies-Recommender-System\Models\user_based_matrix.pkl"
Matrix = load_similarity_matrix(MatrixCSV)
recommendations = get_user_recommendation(DataBase, Matrix,1)
print(recommendations)
|