forked from El-Sayed-Mustafa/Hotel-Rating-Prediction-ML
-
Notifications
You must be signed in to change notification settings - Fork 0
/
phase1.py
189 lines (151 loc) · 7.65 KB
/
phase1.py
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
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
import pandas as pd
import numpy as np
from sklearn.preprocessing import MinMaxScaler
from sklearn.linear_model import LinearRegression
from sklearn.metrics import mean_squared_error as mse
from sklearn.neighbors import KNeighborsRegressor
from sklearn.ensemble import RandomForestRegressor, GradientBoostingRegressor
# Function to winsorize outliers in data
def winsorize_outliers(data, iqr_multiplier=1.5):
q1, q3 = np.percentile(data, [25, 75])
iqr = q3 - q1
upper_whisker = q3 + (iqr_multiplier * iqr)
lower_whisker = q1 - (iqr_multiplier * iqr)
data[data > upper_whisker] = upper_whisker
data[data < lower_whisker] = lower_whisker
return data
# Encoding function for positive reviews
def positiveRev_encoding(value):
exclude_list = ['No Positive', 'Nothing']
for exclude_string in exclude_list:
if exclude_string in value:
return 0
return 1
# Encoding function for negative reviews
def negativeRev_encoding(value):
exclude_list = ['No Negative', 'Nothing', 'Nothing', 'nothing', 'N A']
for exclude_string in exclude_list:
if exclude_string in value:
return 0
return 1
# Extract tags from tags string
def extract_tags(tags_str):
tags = tags_str.strip("[]").replace("'", "").split(", ")
tag_categories = {}
for tag in tags:
if "trip" in tag:
tag_categories["trip_type"] = tag.strip()
elif "Room" in tag:
tag_categories["room_type"] = tag.strip()
elif "Stayed" in tag:
tag_categories["stay_duration"] = tag.strip()
elif "Submitted" in tag:
tag_categories["device_type"] = tag.strip()
else:
tag_categories["group_type"] = tag.strip()
return tag_categories
# Extract country from address
def extract_country(address):
address_parts = address.split()
country = address_parts[-1]
return country
# Preprocess the input DataFrame
def preprocess_data(df):
# Winsorize outliers
columns_to_winsorize = ['Additional_Number_of_Scoring', 'Average_Score', 'Review_Total_Negative_Word_Counts',
'Total_Number_of_Reviews', 'Review_Total_Positive_Word_Counts',
'Total_Number_of_Reviews_Reviewer_Has_Given', 'lat', 'lng', 'Reviewer_Score']
df[columns_to_winsorize] = df[columns_to_winsorize].apply(winsorize_outliers)
# Fill missing values
df['lat'].fillna(df['lat'].mode()[0], inplace=True)
df['lng'].fillna(df['lng'].mode()[0], inplace=True)
# Convert data types
df['days_since_review'] = df['days_since_review'].str.replace("[days]", '').astype(int)
df['Review_Date'] = pd.to_datetime(df['Review_Date'])
# Extract date features
df['month'] = df['Review_Date'].dt.month
df['year'] = df['Review_Date'].dt.year
df['day_of_week'] = df['Review_Date'].dt.day_name()
df['day_of_week'] = df['day_of_week'].map({'Monday': 0, 'Tuesday': 1, 'Wednesday': 2, 'Thursday': 3,
'Friday': 4, 'Saturday': 5, 'Sunday': 6})
# Drop unnecessary columns
df = df.drop(['Review_Date'], axis=1)
# Apply text encoding and cleaning
df['Positive_Review'] = df['Positive_Review'].apply(positiveRev_encoding)
df['Negative_Review'] = df['Negative_Review'].apply(negativeRev_encoding)
df['Hotel_Name'] = df['Hotel_Name'].str.replace('Hôtel', 'Hotel') # Fix typo in hotel names
# Extract tags from 'Tags' column
tags_df = df['Tags'].apply(lambda x: pd.Series(extract_tags(x)))
df = pd.concat([df, tags_df], axis=1)
df.drop('Tags', axis=1, inplace=True)
# Fill missing values and handle categorical variables
df['trip_type'] = df['trip_type'].fillna('Leisure trip')
df['trip_type'] = df['trip_type'].map({'Leisure trip': 1, 'Business trip': 0})
df['room_type'] = df['room_type'].fillna('Double Room')
df['device_type'] = df['device_type'].fillna('Submitted from a mobile device')
df['Hotel_Address'] = df['Hotel_Address'].apply(extract_country)
df['Hotel_Country'] = df['Hotel_Address']
df['Reviewer_Nationality'] = df['Reviewer_Nationality'].str.split().str[-1]
df['Similar_country'] = (df['Hotel_Address'] == df['Reviewer_Nationality']).astype(int)
# Clean and convert 'stay_duration' column
df['stay_duration'] = df['stay_duration'].str.replace("[Stayed, night,s]", '')
mode_duration = df['stay_duration'].mode()[0]
df['stay_duration'] = df['stay_duration'].fillna(mode_duration).astype(int)
df['stay_duration'] = winsorize_outliers(df['stay_duration'])
return df
# Read the CSV file and shuffle the DataFrame
df = pd.read_csv("hotel-regression-dataset.csv")
df_shuffled = df.sample(n=len(df), random_state=1)
# Preprocess the shuffled DataFrame
df_shuffled = preprocess_data(df_shuffled)
# Select relevant columns for the final DataFrame
df_final = df_shuffled[
['Additional_Number_of_Scoring', 'Average_Score', 'Negative_Review', 'Review_Total_Negative_Word_Counts',
'Total_Number_of_Reviews', 'Positive_Review', 'Review_Total_Positive_Word_Counts',
'Total_Number_of_Reviews_Reviewer_Has_Given', 'days_since_review', 'lat', 'lng', 'month', 'year', 'day_of_week',
'trip_type', 'stay_duration', 'Similar_country', 'Reviewer_Score']]
# Split the final DataFrame into train, validation, and test sets
train_df, val_df, test_df = df_final[:174189], df_final[174189:232252], df_final[232252:]
# Split the features and target variables
X_train, y_train = train_df.iloc[:, :-1].values, train_df.iloc[:, -1].values
X_val, y_val = val_df.iloc[:, :-1].values, val_df.iloc[:, -1].values
X_test, y_test = test_df.iloc[:, :-1].values, test_df.iloc[:, -1].values
# Scale the numerical features using MinMaxScaler
scaler = MinMaxScaler().fit(X_train[:, :8])
def preprocessor(X):
A = np.copy(X)
A[:, :8] = scaler.transform(A[:, :8])
return A
# Preprocess the features
X_train, X_val, X_test = preprocessor(X_train), preprocessor(X_val), preprocessor(X_test)
# Train and evaluate Linear Regression model
lm = LinearRegression().fit(X_train, y_train)
train_mse_lm = mse(lm.predict(X_train), y_train)
val_mse_lm = mse(lm.predict(X_val), y_val)
# Train and evaluate K-Nearest Neighbors model
knn = KNeighborsRegressor(n_neighbors=14).fit(X_train, y_train)
train_mse_knn = mse(knn.predict(X_train), y_train)
val_mse_knn = mse(knn.predict(X_val), y_val)
# Train and evaluate Random Forest model
rfr = RandomForestRegressor(max_depth=10).fit(X_train, y_train)
train_mse_rfr = mse(rfr.predict(X_train), y_train)
val_mse_rfr = mse(rfr.predict(X_val), y_val)
# Train and evaluate Gradient Boosting model
gbr = GradientBoostingRegressor(n_estimators=100).fit(X_train, y_train)
train_mse_gbr = mse(gbr.predict(X_train), y_train)
val_mse_gbr = mse(gbr.predict(X_val), y_val)
# Calculate MSE for the Gradient Boosting model predictions on the test set
gbr_pred = gbr.predict(X_test)
test_mse_gbr = mse(gbr_pred, y_test)
# Calculate the score for the Gradient Boosting model on the test set
score_gbr = gbr.score(X_test, y_test)
print("Train MSE (Linear Regression):", train_mse_lm)
print("Validation MSE (Linear Regression):", val_mse_lm)
print("Train MSE (K-Nearest Neighbors):", train_mse_knn)
print("Validation MSE (K-Nearest Neighbors):", val_mse_knn)
print("Train MSE (Random Forest):", train_mse_rfr)
print("Validation MSE (Random Forest):", val_mse_rfr)
print("Train MSE (Gradient Boosting):", train_mse_gbr)
print("Validation MSE (Gradient Boosting):", val_mse_gbr)
print("Test MSE (Gradient Boosting):", test_mse_gbr)
print("Gradient Boosting Score:", score_gbr)