Secure water quality prediction system using machine learning and blockchain technologies.

Area of Science:

• Environmental Science
• Computer Science
• Data Science

Background:

• Freshwater is a scarce resource, with only 2-3% available globally.
• Pollution from sewage, waste, and industrial effluents threatens freshwater bodies.
• Maintaining water quality is crucial for human survival and ecosystem health.

Purpose of the Study:

• To develop a secure and accurate system for predicting water quality parameters.
• To evaluate the performance of various machine learning models for water quality prediction.
• To integrate blockchain technology for enhanced security in water quality data management.

Main Methods:

• Machine learning models including Linear Regression, Generalized Linear Model, Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), classification and regression trees, and Random Forest were employed.
• Geographical coordinates (latitude and longitude) were used as input for prediction.
• Blockchain technology, Secure Hash Algorithm-256 (SHA), and Rivest-Shamir-Adleman (RSA) were utilized for security and data transfer.

Main Results:

• The Random Forest (RF) algorithm demonstrated superior prediction accuracy compared to other models.
• RF achieved a mean absolute error of 0.56, mean square error of 0.33, and root mean square error of 0.56.
• The integration of blockchain ensured a secure prediction process with verified authorized personnel.

Conclusions:

• Machine learning, particularly Random Forest, offers a robust solution for water quality prediction.
• Blockchain technology provides a secure framework for managing and verifying water quality data.
• This integrated system can contribute to reducing pollution and improving overall water quality.