The goal of this project is to complete a study on the different exisiting methods that could be implemented in order to predict the future CPU usage of servers : Time Series Forcasting (TSF).
The study focuses on two main categories of algorithms : statistical methods and machine learning methods.
In this repo, you will find all the code that I implemented, as well as the different datasets used, in order to complete this study and learn more about TSF.
3 different datasets were used in this study :
- Azure CPU Usage : https://github.com/amcs1729/Predicting-cloud-CPU-usage-on-Azure-data
- BitBrains dataset CPU Usage : http://gwa.ewi.tudelft.nl/datasets/gwa-t-12-bitbrains
- KSA ministry of finance dataset : https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6138836/
In the files above, the "final_notebook.ipynb" is the final output of this study, where all the results can be seen. The notebook has been built in order to be run as a voila notebook and used as an interface to showcase the results during a presentation