https://github.com/funinkina/water-usability-analysis

This repository containes the files for the data and code used to build a neural network to determine the usablity of any water sample in dairy industry,
https://github.com/funinkina/water-usability-analysis

industrial neural-network water-quality

Last synced: 8 months ago
JSON representation

This repository containes the files for the data and code used to build a neural network to determine the usablity of any water sample in dairy industry,

• Host: GitHub
• URL: https://github.com/funinkina/water-usability-analysis
• Owner: funinkina
• Created: 2024-07-24T17:39:13.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2024-07-24T17:42:11.000Z (about 1 year ago)
• Last Synced: 2024-07-24T20:27:35.006Z (about 1 year ago)
• Topics: industrial, neural-network, water-quality
• Language: Jupyter Notebook
• Homepage:
• Size: 3.29 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
In this study, we present a machine learning-based approach to

predict the usability of water for the dairy industry based on its various

physicochemical parameters. The dairy industry requires water of specific

quality to ensure product safety and operational efficiency. We utilized a

dataset comprising parameters such as Turbidity (NTU), pH, Conductivity

(uS/cm), Total Dissolved Solids (mg/l), Total Hardness (mg/l as CaCO3),

Aluminium (mg/l), Chloride (mg/l), Total Iron (mg/l), Sodium (mg/l),

Sulphate (mg/l), Biochemical Oxygen Demand/Zinc (mg/l), Magnesium

(mg/l), Calcium (mg/l), Nitrate (mg/l), and Phosphate (mg/l). These

parameters were used to train a neural network model to predict water

usability, classified into three categories: not usable, conditionally usable,

and fully usable.  

The model was developed using a synthetic dataset, with

preprocessing steps including normalization and handling missing values.

Feature importance was assessed using machine learning techniques to

understand the significance of each parameter in determining water

quality for dairy applications. The trained model demonstrated high

accuracy and reliability in predicting water usability, offering a valuable

tool for the dairy industry to assess and manage water resources

efficiently.   

With the help of compiled model by tensorflow, we can use it to

determine the usage of the water on a scale of 0, 1 or 2. 0 meaning

competely unusable. 1 meaning only non important parameters are

unusable, and 2 means the water sample is perfectly usable.

----

The dataset is included in the repository as well as thorough description in the pdf.