A Methodology Based on Machine Learning and Soft Computing to Design More Sustainable Agriculture Systems

Area of Science:

• Agricultural Technology
• Data Science
• Artificial Intelligence

Background:

• Technological advancements like IoT, cloud computing, and machine learning offer powerful tools for decision support.
• The agricultural sector faces challenges in sustainability, necessitating intelligent solutions for improved decision-making.

Purpose of the Study:

• To propose a methodology for developing Decision Support Systems (DSS) in agriculture using IoT data and Soft Computing.
• To enhance farmer decision-making capabilities for improved agricultural sustainability.

Main Methods:

• Utilizing time-series data from the Internet of Things (IoT) ecosystem.
• Applying data preprocessing and modeling with machine learning techniques within the Soft Computing framework.
• Developing a predictive model for a defined future time horizon.

Main Results:

• The methodology generates a predictive model capable of inference for a given prediction horizon.
• The system was applied to early frost prediction, demonstrating its practical utility.
• Validation with expert farmers confirmed the methodology's effectiveness and benefits.

Conclusions:

• The proposed methodology effectively integrates IoT, machine learning, and Soft Computing for agricultural DSS.
• The system provides valuable predictive insights, such as early frost warnings, to support farmers.
• The approach enhances agricultural sustainability through intelligent data-driven decision-making.