Optimising deep learning models for prediction of dry matter content in mango fruit with an extensive multi-population, multi-instrument spectral dataset
This repository accompanies Chapter 4 in the submitted thesis of Jeremy Walsh, 2024, Central Queensland University, "Deep Learning In Estimation of Fruit Attributes Using Near Infrared Spectroscopy".
The code within this repository was as described in the methodology in the aforementioned thesis chapter. Computational analyses and modelling tasks were carried out on a laptop equipped with a 12th Gen Intel® Core™ i7-1265U processor, 32 GB of RAM, and a shared Intel® Iris® Xe Graphics GPU, running Windows 10 Enterprise. Optuna optimisation studies were executed on a separate workstation featuring a 12th Gen Intel® Core™ i9-12950HX processor, 64 GB of RAM, and a dedicated NVIDIA GeForce RTX 3080 GPU, operating on Ubuntu 20.04.6 LTS. The software environment for both setups was based on Python 3.10, and the requirements.txt file specifies all necessary Python packages and libraries, along with their respective versions.
The extensive mango dry matter and near-infrared spectra dataset used in this study is published on Mendeley Data, accessible at: data.mendeley.com/datasets/46htwnp833/4.
Should you require additional information or have any questions regarding this study, please do not hesitate to get in touch through the following channels:
Email: [email protected]
LinkedIn: Jeremy Walsh