Data mining derived from food analyses using non-invasive/non-destructive analytical techniques; determination of food authenticity, quality & safety in tandem with computer science disciplines

• Monitoring/predicting quality and safety of foods relies on costly, time-consuming conventional microbiological analyses.
• Advances in sensor technology, signal and data analysis could help overcome these limitations.
• Data from sensor devices are analyzed with chemometric, machine learning and computational intelligence techniques.
• The developed rapid methods can be applied on-, in- or at-line in the context of Process Analytical Technology.

BackgroundFood quality, safety and authenticity are important issues for consumers, governments, as well as the food industry. In the last decade, several researchers have attempted to go beyond traditional microbiological, DNA-based and other methods using rapid techniques. This broad term involves a variety of sensors such as hyperspectral and multispectral imaging, vibrational spectroscopy, as well as biomimetic receptors.Scope and approachThe resulting data acquired from the above-mentioned sensors require the application of various case-specific data analysis methods for the purpose of simple understanding and visualization of the acquired high-dimensional dataset, but also for classification and prediction purposes.Key findings and conclusionsIt is evident that rapid techniques coupled with data analysis methods have given promising results in several food products with various sensors. Additionally there are several applications, new sensors and new algorithms that remain to be explored and validated in the future.