The wine industry has been striving to achieve differentiation in their products and to improve their in- trinsic quality and consistency. Ensuring quality involves harvesting grapes at the optimal maturity point and selecting them according to the desired characteristics of the wine to be produced.
In this context, hyper- spectral imaging (HSI) combined with machine learning algorithms is a promising alternative to predict important enological parameters and assist on harvesting critical decisions.
However, the large amount of data generated by HSI includes not only relevant but also a lot of redundant information that raise computational challenges for data-driven modelling. Several machine learning approaches have been proposed to handle such data characteristics, but selecting a suitable method is a cumbersome task.
In this work, a predictive analytics comparison framework (PAC) was applied to estimate relevant enological parameters for grape ripeness assessment.