BlendedCrudeDistillationProfile

Using data science methods to build a model to predict the distillation profile of blended crude oils.

This project consists of web crawling, data cleaning, model selection, training, and hyperparameter tuning, and finally deployment using Flask. A little bit of HTML was used to create a simple user interface.

Files:

• WebCrawler.py: The Python module responsible for performing web scrapping.
• DataCleaner.py: The Python module used to clean the web data and transform them to be ready for modelling.
• CrudeBlendModel.py: Main logic for blending rules and machine learning models.
• main.py: The file that automates every step and creates a simple user interface.
• templates: The directory which contains some HTML files for the UI.
• solution_summary.ipynb: A summary of the solution I used to solve this project.
• solution_summary.pdf: A pdf version of solution_summary.ipynb
• TestDataCleaner.py: Unit test for the module DataCleaner.py. Includes a couple simple test cases.

To run this program on your local computer, clone this directory, and run main.py using Python. It will take a few minutes for the server to get started. The web application will run on your localhost.