statisticianinstilettos / recommendation_system_creditcard

Recommendation_System_Using_CreditCard_Transactional_Data

Create recommendation system based on credit card transactional data.

Project Organization

├── LICENSE
├── README.md          <- The top-level README for developers using this project.
│
├── notebooks          <- Jupyter notebooks.
│
│
├── reports            <- Generated analysis in PDF
│   └── figures        <- Generated graphics and figures to be used in reporting
│
├── requirements.txt   <- The requirements file for reproducing the analysis environment, e.g. generated with `pip freeze > requirements.txt

---

Goals:

Create recommendation system(s) using credit card transactional records. The data contains transactions from multiple users and contains GPS coordinates. The goals of the recommendation system(s) are as follows:

Goal #1.The recommendation system should recommend similar merchants to the customer based on similarities between the merchants

Goal #2.The recommendation system should recommend merchants to the customer based on their current location.

Notebooks:

Along with EDA.ipyb and pre-work for goal #2 notebooks (Recommendation_System_Location_Based_Pre_Work.ipynb) following notebooks capture the code for the two goals.

----A notebook titled "Recommendation_System_v1.ipynb" captures code for goal #1.

----A notebook titled "Recommendation_System_Location_Based_HDBSCAN.ipynb" captures code for goal #2.

(Note: A notebook titled "Recommendation_System_Location_Based_Kmeans.ipynb" captures code for goal #2 however it is only for learning purposes)

Reports:

Documentation on the project (project presentation, milestone reports etc.) are provided in reports folder. This folder also contains figures used to create the reports.