Title : Hotel-Booking-Predicting-System

About the Dataset:

• Hotel booking demand datasets
• A data set has information about bookings for two hotels, a city hotel and a resort hotel. The information includes when the booking was made, how long the stay was, how many people were staying, and if parking was available.
• Source: Dataset taken from kaggle website

Description

The data set has one file that compares information about bookings for a city hotel and a resort hotel.

Attributes: Data columns (total 32 columns):
No. Column
0. hotel

1. is_canceled
2. lead_time
3. arrival_date_year
4. arrival_date_month
5. arrival_date_week_number
6. arrival_date_day_of_month
7. stays_in_weekend_nights
8. stays_in_week_nights
9. adults
10. children
11. babies
12. meal
13. country
14. market_segment
15. distribution_channel
16. is_repeated_guest
17. previous_cancellations
18. previous_bookings_not_canceled
19. reserved_room_type
20. assigned_room_type
21. booking_changes
22. deposit_type
23. agent
24. company
25. days_in_waiting_list
26. customer_type
27. adr
28. required_car_parking_spaces
29. total_of_special_requests
30. reservation_status
31. reservation_status_date

Getting Started

Business Questions:

List of Questions to help project goals

1. How Market Segment Of Booking Affecting Cancellation ?
2. How long do people stay at the hotels?
3. Which are the most busy months?
4. What Are The Other Factors that affecting cancellation of booking ?
5. Which countries do customers come from?
6. What types of customers are most common in each hotel?

What machine learning algorithm that has the highest accuracy when it comes predicting hotel booking cancellations ?

Workflow:

1. Data Cleaning :

Imputing missing value with mean Dropping rows with abnormal values: 0 Total guests / adults in the booking

2. Exploratory Data Analysis :

Feature Engineering Aggregating Columns - the agg function refers to the aggregation operation that is being performed on the data. Visualization Insight & Conclusion

3. Feature Selection for machine learning process

Label encoding for certain columns that needs to be encoded

4. Model Building

• Train Test Split
• Using pipeline for model building
* scaling for numerical features
* Normalizing for numerical features
• • Creating base model with few algorithm
    * Logistic Regression,
    * K Neighbors Classifier,
    * Decision Tree Classifier,
    * Random Forest classifier
• Checking evaluation matrix
• Comparing the model with the best accuracy score

Authors

@Kaaviasudhan

Version History

• 0.1
  • Initial Release