alhudareza4 / Project-Hotel-Demand-Classification

Project-Hotel-Demand-Classification

Created By : Alhuda Reza Mahara

Description

This notebook provides an end-to-end demonstration of Hotel Demand analytics: Booking Cancellation Classification using machine learning. The scope of topics discussed in this notebook includes data analysis & machine learning (supervised classification, imbalanced dataset problem, hyperparameter tuning, etc.)

This notebook also serves as a part of graduation requirement from Job Connector Data Science program by Purwadhika.

Problem Statement

• Kita berupaya membantu pihak pemilik hotel untuk melakukan klasifikasi pembatalan pesanan Hotel.
• Kita ingin mengetahui penyebab customer membatalkan pesanan Hotel, sehingga pihak hotel dapat mencegah penyebab pembatalan Hotel.
• Pihak Hotel membutuhkan model untuk dapat mengklasifikasikan customer sehingga mereka dapat melakukan tindakan preventif untuk mencegah customer melakukan pembatalan. Untuk menjawab pertanyaan :

1. Matriks apa yang digunakan ?
2. Model apa yang paling cocok digunakan untuk model ini ?
3. Bagaimana performa model ?
4. Kesimpulan dan Rekomendasi untuk pihak hotel ?

Dataset Informaton

We're using Hotel Boking demand dataset. The dataset can be found in this folder.

Contents

1. Purpose

2. Business Problem Understanding

3. Library & Function

3.1 Python Package Library

3.2 Defined Function

4. Data Understanding

4.1 Attribute Information

4.2 Data Understanding

  4.2.1 Missing Values

  4.2.2 Missing Values handling

  4.2.3 Target & Feature Check
  
    4.2.3.1 Target Check
  
    4.2.3.2 Feature Check

4.3 Feature Correlation

  4.3.1 Numerical Feature With Target

  4.3.2 Categorical Feature With Target

  4.3.3 Multiple Feature With Target

5. Machine Learning

5.1 Data Preparation

5.2 Base Model Evaluation

5.3 Handling Imbalance

  5.3.1 Base Model with Oversampling (Random Over Sampling)

  5.3.2 Base Model with Undersampling (Random Under Sampling)

  5.3.3 Base Model with Undersampling (SMOTE)

  5.3.4 Base Model with Undersampling and Oversampling (SMOTEtomek)

5.4 Recap Modul Output

  5.4.1 Recap Model Output : Train Data 

  5.4.2 Recap Model Output : Test Data

5.5 Hyperparameter Tuning

5.6 Feature Importance

6. Conclusion

7. Recomendation

8. Deployment

How To Use

To get the notebook, simply download or clone this repository. You can use VS Code, Jupyter Notebook, or others of your choice