On Agriculture, Forest and Environmental Geodata Statistical Analysis, Modelling and Machine Learning [SAFEST]

Overview

The On Agriculture, Forest and Environmental Geodata Statistical Analysis, Modelling and Machine Learning [SAFEST] will provide theoretical and practical teaching on the statistical methods and tools for geodata handling and analysis, with special emphasis to soil traits, crop yield or plant biomass in the agricultural, forest and other land uses under variable management and environmental conditions. Methods provided will be tailored to a wide range of scales, from plot to landscape to regional scales.

Methods provided will be tailored to a wide range of scales, from plot to landscape to regional scales.

Topics in the school include:

• advanced literature search and meta-analysis;
• use of reference databases on land cover, soil traits, and meteorological and climate data;
• data visualization, spatial references and projections, proximal and remote sensed data, terrain analysis;
• methods for covariate identification, acquisition, and selection, harmonization, and inclusion in modelling procedures;
• linear mixed models for statistical analysis of soil and biological data, methods to include spatialization and soil depth as variables, methods to study unbalanced data or including variables with missing data, regression models;
• overview on models and some case studies: classification and regression trees models (random forest, boosted regression trees, others), artificial neural networks, and convolutional neural networks, etc..

The school will include theoretical lessons in the morning and practical lessons in the afternoon.

Participants to SAFEST will be able to deal with geodata and their handling and analysis in a wealth of areas, including agriculture, forest, other land uses, the study of soil biodiversity, soil science, farming systems, and other environmental areas.

The Summer School received the support of SHARInG-MeD project (a PRIMA 2022 action topic 1.2.1 under the grant agreement 2211).