This project was made in collaboration with Eya Mlika
https://drive.google.com/file/d/1JqxixnxOGOYmO8ALl7TX1Oy9yXabLGvo/view?usp=share_link
This repository contains all of our activities realized during our P2M project at Sup’com. Our mission was to develop a smart layer to an existing Iot solution "cool@-C". In order to make better decisions and save energy. The work consists of applying machine learning algorithms to predict the right temperature and to detect abnormal cases related to the activities of the air conditioner, consequentially managing the energy costs of an air conditioner. This project had has as a goal to exploit our knowledge acquired in class and make us gain more in depth practical training.This project allowed us to become familiar with the machine learning algorithms and to discover Time series algorithms, learn a new web development framework and get acquainted with a NoSql SGBD.
We have structured our report in 6 parts, starting with an introduction and the definition of the state of the art. Then we delve deeper into the project analysis, the
methodologies and technological choices we made in development. In the model implementation section,which follows the CRISP-DM methodology for data mining, we talk more about the data and machine learning algorithms we used. Followed by the web application section where we talk in depth of how we developed and exploited the results found in the previous section.
Finally, we will end with a conclusion, in which we will summarize all that has been achieved. We will highlight the added values that we have brought to this project. This project can be subject to evolutionary extensions, in the framework of continuous improvement; this will be the subject of the perspectives