Operationally Realistic Validation for Prediction of Cocoa Sensory Qualities by High-Throughput Mass Spectrometry

Abstract

The potential of analytical chemistry to predict sensory qualities of food materials is a major current theme. Standard practice is cross-validation (CV), where a set of chemical and associated sensory data is partitioned so chemometric models can be developed on training subsets, and validated on held-out subsets. CV demonstrates prediction, but is an unlikely scenario for industrial operations, where concomitant data acquisition for model development and test materials would be unwieldy. We evaluated cocoa materials of diverse provenance, and analyzed on different dates to those used in model development. Liquor extracts were analyzed by flow-injection electrospray-mass spectrometry (FIE-MS), a novel method for sensory quality prediction. FIE-MS enabled prediction of sensory qualities described by trained human panelists. Optimal models came from the Weka data-mining algorithm SimpleLinearRegression, which learns a model for the attribute giving minimal training error, which was (-)-epicatechin. This flavonoid likewise dominated partial least-squares (PLS)-regression models. Refinements of PLS (orthogonal-PLS or orthogonal signal correction) gave poorer generalization to different test sets, as did support vector machines, whose hyperparameters could not be optimized in training to avoid overfitting. In conclusion, if chemometric overfitting is avoided, chemical analysis can predict sensory qualities of food materials under operationally realistic conditions.

Validation of chemometric models – A tutorial

“…[13] it is pointed out that "cross-validation demonstrates prediction, but is an unlikely scenario in industrial applications, where concomitant data acquisition for model development and test materials would be unwieldy". In this context, the same applies to random splits into calibration and test sets.…”

Merits of random forests emerge in evaluation of chemometric classifiers by external validation

“…Capture of real-world data scenarios in chemometrics papers has been rare, due to limitations of sample availability [1,7]. This has caused disputes over the merits of compromises for validation, with different fields offering divergent perspectives [5,8].…”

Sum of ranking differences (SRD) to ensemble multivariate calibration model merits for tuning parameter selection and comparing calibration methods

“…Different data sets were used for the comparison. In other recent studies, SRD was used to compare rankings of sensory models relative to panel scores [53,54], different curve resolution and classification methods were compared using a variety of performance merits [55,56]. Lastly, among the diverse applications, SRD has been used to compare several modeling methods to compare and form quantitative structure activity relationship (QSAR) models [34,57].…”