Abstract
The dissertation showcases the effectiveness of explainable machine learning approaches in studying plant ecophysiology and agriculture. It demonstrates the identification and prioritization of ecologically relevant traits using such methods in the genus Helianthus (wild sunflowers). Phenotypic differentiation and interspecific diversification are explored, as well as intraspecific trait variations within Helianthus annuus across different ecological regions. Additionally, the dissertation applies similar methods to assess the impact of historical weather patterns on the agricultural yield of cultivated sunflower at national and regional scales. It also provides yield forecasts under future socioeconomic scenarios, considering the potential effects of climate change on sunflower cultivation. Overall, this work highlights the potential of machine learning coupled with interpretable methods, in analyzing nonlinear and multidimensional biological data, addressing important research questions in plant biology, ecology, and agriculture. The findings contribute to understanding evolutionary predictability, ecological strategies, and the impact of climate change on crop yields.
Notes
If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at STARS@ucf.edu
Graduation Date
2023
Semester
Summer
Advisor
Mason, Chase
Degree
Doctor of Philosophy (Ph.D.)
College
College of Sciences
Department
Biology
Degree Program
Integrative Conservation Biology; Integrative Biology
Identifier
CFE0009752; DP0027860
URL
https://purls.library.ucf.edu/go/DP0027860
Language
English
Release Date
August 2024
Length of Campus-only Access
1 year
Access Status
Doctoral Dissertation (Campus-only Access)
STARS Citation
Majumder, Sambadi, "Applications of Interpretable Machine Learning Methods in Plant Ecology and Crop Science" (2023). Electronic Theses and Dissertations, 2020-2023. 1802.
https://stars.library.ucf.edu/etd2020/1802
Restricted to the UCF community until August 2024; it will then be open access.