Mapping of soil suitability for medicinal plants using machine learning methods.

Area of Science:

• Integrative biology and computational science
• Environmental science and conservation
• Pharmacognosy and drug discovery

Background:

• Medicinal plant conservation is vital but challenged by climate change, over-harvesting, and habitat loss.
• Traditional assessment methods are time-consuming and prone to errors.
• Sustainable utilization of medicinal herbs requires innovative conservation strategies.

Purpose of the Study:

• To develop a swift decision-making approach for conserving vulnerable medicinal plants.
• To enhance the productivity of curative plants using data-driven methods.
• To promote drug discovery through improved conservation and identification of medicinal herbs.

Main Methods:

• Utilized machine learning algorithms (Extra Tree Classifier, Random Forest, etc.) and Geographic Information Systems (GIS).
• Developed a novel dataset using a quantum GIS tool for spatial analysis.
• Implemented supervised algorithms for classifying soil types and subregions relevant to medicinal plants.

Main Results:

• The Extra Tree Classifier (EXTC) model achieved high accuracy (99.01% for soil, 98.76% for subregion classification).
• The system can predict suitable soil classes for a given subregion and vice versa.
• Choropleth maps visualize potential medicinal herbs and their conservation status based on soil and subregion.

Conclusions:

• The developed approach offers a comprehensive and swift reference for conserving medicinal herbs.
• Machine learning and GIS integration provide effective tools for spatial analysis and conservation planning.
• This method supports conservationists, researchers, and the public in identifying and protecting vulnerable medicinal plants.