54 lines
2.5 KiB
Python
54 lines
2.5 KiB
Python
|
import math
|
||
|
|
|
||
|
|
from sklearn.neural_network import MLPRegressor
|
||
|
|
from sklearn.model_selection import train_test_split
|
||
|
|
from sklearn.preprocessing import StandardScaler, LabelEncoder
|
||
|
|
from sklearn.metrics import mean_squared_error, mean_absolute_error, r2_score, accuracy_score, classification_report
|
||
|
|
import pandas as pd
|
||
|
|
|
||
|
|
# Загрузим данные
|
||
|
|
df = pd.read_csv("hotel_bookings_raw.csv")
|
||
|
|
df.dropna(inplace=True)
|
||
|
|
|
||
|
|
# Объявляю объект для преобразования строковых значений в числовые
|
||
|
|
label_encoder = LabelEncoder()
|
||
|
|
|
||
|
|
# Выберем признаки и целевую переменную (доход)
|
||
|
|
features_list = ['lead_time', 'stays_in_weekend_nights',
|
||
|
|
'stays_in_week_nights', 'adults', 'children', 'babies', 'meal', 'customer_type',
|
||
|
|
'previous_cancellations',
|
||
|
|
'previous_bookings_not_canceled', 'required_car_parking_spaces',
|
||
|
|
'CPI_AVG', 'INFLATION', 'INFLATION_CHG', 'GDP', 'CPI_HOTELS']
|
||
|
|
|
||
|
|
# Выбираем признаки и целевую переменную
|
||
|
|
features = df[features_list].copy()
|
||
|
|
target = df['adr'].copy()
|
||
|
|
|
||
|
|
# Применяю к каждому столбцу признака преобразования
|
||
|
|
for f in features_list:
|
||
|
|
features[f] = label_encoder.fit_transform(features[f])
|
||
|
|
|
||
|
|
# Разделяем данные на тренировочный и тестовый наборы
|
||
|
|
X_train, X_test, y_train, y_test = train_test_split(features, target, test_size=0.1, random_state=42)
|
||
|
|
|
||
|
|
# Масштабируем признаки для лучшей производительности модели
|
||
|
|
scaler = StandardScaler()
|
||
|
|
X_train_scaled = scaler.fit_transform(X_train)
|
||
|
|
X_test_scaled = scaler.transform(X_test)
|
||
|
|
|
||
|
|
# Создаем и обучаем MLPRegressor
|
||
|
|
mlp_model = MLPRegressor(hidden_layer_sizes=(100, 50), max_iter=500, random_state=42)
|
||
|
|
mlp_model.fit(X_train_scaled, y_train)
|
||
|
|
|
||
|
|
# Делаем предсказания на тестовых данных
|
||
|
|
predictions = mlp_model.predict(X_test_scaled)
|
||
|
|
|
||
|
|
# Оцениваем производительность модели
|
||
|
|
mse = mean_squared_error(y_test, predictions)
|
||
|
|
mae = mean_absolute_error(y_test, predictions)
|
||
|
|
r2 = r2_score(y_test, predictions)
|
||
|
|
|
||
|
|
print(f"Среднеквадратичная ошибка (MSE): {round(math.sqrt(mse), 2)}%")
|
||
|
|
print(f"Среднеабсолютное отклонение (MAE): {round(mae, 2)}%")
|
||
|
|
print(f"Коэффициент детерминации (R^2): {round(r2, 4) * 100}%")
|