Skip to main content
  • Home
  • Thesis Topics 
  • Precision Agriculture with Machine Learning: Influence of Soil Fertility on Coffee Yield

Research Thesis Topic

Precision Agriculture with Machine Learning: Influence of Soil Fertility on Coffee Yield


Topic ID
325

Thesis Topic/Title
Precision Agriculture with Machine Learning: Influence of Soil Fertility on Coffee Yield

Description

Coffee is the second most important traded commodity in African, American and Asian countries. Yield is significantly dependent on environmental, climatic and soil fertility conditions, and especially the soil fertility constituents whose correct proportions can impact the productivity. As it is not known what soil fertility conditions are the best for optimum yield, this remains a challenging task for agricultural modellers.

Biophysical models require a knowledge of the processes in soil-plant-atmosphere continuum that affects optimal crop yield. Owing to the complexity and parameterisation of best fertility properties for crop simulation models, statistical analytics built in an intelligent model can provide a new a predictive approach for yield optimisation through carefully selected soil fertility properties. The soil organic matter containing most of the soil’s reserve of nitrogen and large portions of phosphorus and sulphur plays a vital role for nutrient availability and absorption by plants but when the process of selecting and reducing variables is automated, predictors chosen might not highlight the biological assumptions.

The purpose of this project is to investigate robust data-intelligent methods [i.e., extreme machine learning (ELM) and random forest (RF)] to model the relationships between soil fertility parameters and coffee yield for smallholder farms. The project will determine the best performing model and relevant organic matter, available potassium and sulphur, etc as the predictors of coffee yield. The project is scalable for further research, such as to a Masters or a PhD program to advance research careers, acquire programming skills and modelling.

The project suits students with a background in agriculture, environmental science, mathematics, computing and climate/atmospheric physics. Students will learn about neural network models, machine learning in agriculture, expert systems with smart tools and evaluation of models in real-time forecasting crop yield.


Principal Supervisor

Associate Supervisors

Research Affiliations
  • Institute for Agriculture and the Environment
  • International Centre for Applied Climate Sciences
  • School of Agricultural, Computational and Environmental Sciences

Field of Research
  • Agricultural Biotechnology
  • Agriculture, Land and Farm Management
  • Atmospheric Sciences
  • Ecological Applications
  • Environmental Science and Management
  • Other Earth Sciences
  • Other Environmental Sciences


Application Open Date
11/02/2018

Application Close Date
31/12/2020

USQ Scholarship Applications

Other Scholarship Funding Details
Australia Awards (Country-Specific Closing Dates), and Australian Endeavor Scholarships (Close June 2018)

Pre-approved for Ethics
Not Applicable

Admission Requirements

Please review the admission requirements for the academic program associated with this Thesis Topic

The project suits students with a background in agriculture, environmental science, mathematics, computing and climate/atmospheric physics. Students will learn about neural network models, machine learning in agriculture, expert systems with smart tools and evaluation of models in real-time forecasting crop yield.




Back to List