Intelligent analysis of maleic hydrazide using a simple electrochemical sensor coupled with machine learning

Area of Science:

• Electrochemistry
• Materials Science
• Data Science

Background:

• Maleic hydrazide (MH) is a plant growth regulator with potential health implications.
• Accurate detection of MH residues in agro-products is crucial for food safety.
• Existing detection methods can be complex and costly.

Purpose of the Study:

• To design a simple, low-cost electrochemical sensing platform for MH detection.
• To integrate machine learning (ML) for intelligent analysis of MH in food matrices.
• To develop a flexible, disposable electrode for practical agro-product safety applications.

Main Methods:

• Fabrication of a laser-induced porous graphene (LIPG) flexible electrode via direct laser writing.
• Development of ML models including artificial neural networks (ANN), random forest (RF), and least squares support vector machine (LS-SVM).
• Optimization of ML models using a data partitioning technique for optimal concentration ranges and dataset sizes.

Main Results:

• The LIPG electrode exhibited excellent flexibility, a 3D porous structure, large surface area, and good conductivity.
• The LS-SVM model demonstrated superior performance compared to ANN and RF for MH detection.
• The developed sensing platform achieved high recovery rates and low relative standard deviations (RSD) in practical applications.

Conclusions:

• A cost-effective and flexible electrochemical sensing platform for MH detection was successfully developed.
• The integration of LS-SVM provides an intelligent and accurate method for analyzing harmful residues in agro-products.
• This approach holds significant potential for enhancing food safety monitoring and intelligent analysis systems.