System that predicts the origin of fecal matter contamination in water.
Fecal matter contains bacteria that in contact with other organisms such as humans, can be harmful. Drinking water contaminated by fecal matter can cause stomach and intestinal illness including diarrhea and nausea, and even lead to death.
In order to prevent human exposition to tainted water and to detect the causative source, microbial source tracking (MST) describes a suite of methods and an investigative strategy for determination of fecal pollution sources in environmental waters that rely on the association of certain fecal microorganisms with a particular host. These values depend on the source, dissolution, time, localization and season. Retrieving them is an expensive process.
In this project, we keep with the dynamic of recent literature where machine learning methods have been used to successfully detect the causative source. To contribute to this task, an online platform is created. This platform allows the user to combine microbiological analysis of water with machine learning techniques to detect the source of contamination and allows the user to introduce microbiological information, visualize it, create and execute prediction models and interact with the obtained results.
This project will allow the scientist community applying machine learning algorithms into MST data isolating them from difficulties such as the environment, programming languages and the economic cost.
Slides with an explanation of the overall work done as final degree project:
- https://docs.google.com/presentation/d/1-SwO1AOSW79MwHgpcJ2nX0QHzEpIv4e_Bg78l89PSi8/edit?usp=sharing
- You can also find the whole memory in this repo.