Predicting Ecotoxicological Effects of pesticides on Honeybee (Apis mellifera) Using Machine Learning

Description

This dataset accompanies the manuscript "Machine learning to extrapolate honeybee chronic ecotoxicity for data-poor pesticides" submitted to Ecotoxicology and Environmental Safety. It contains all input data files and the Python source code used to develop, evaluate, and deploy Random Forest and Artificial Neural Network models predicting honeybee chronic ED10 values for adult and larval life stages, together with a tri-tiered K-Nearest Neighbor applicability domain framework for quantifying prediction uncertainty across over 300 uncharacterized pesticides. The script is research code intended for transparency and peer review verification

Institutions

• Technical University of Denmark

  Capital Region, Kongens Lyngby

Categories

Licence