Evaluation of Coffee Roasting Degree by Using Electronic Nose and Artificial Neural Network for Off‐line Quality Control

Abstract

Actually the evaluation of coffee roasting degree is mainly a manned operation, substantially based on the empirical final color observation. For this reason it requires well-trained operators with a long professional skill. The coupling of e-nose and artificial neural networks (ANNs) may represent an effective possibility to roasting process automation and to set up a more reproducible procedure for final coffee bean quality characterization.

Application of multivariate analysis and mass transfer principles for refinement of a 3-L bioreactor scale-down model-when shake flasks mimic 15,000-L bioreactors better

“…In addition to projection‐based methodologies, other chemometric techniques like support vector machines can also be used, which allow modeling the nonlinear features of the data using kernels . Other chemometric techniques including various neural network methodologies may also be used …”

“…49 Other chemometric techniques including various neural network methodologies may also be used. 50,51 There does not appear to be much agreement within the scientific community regarding the approach that should be taken to qualify a scale-down model for a cell culture process. Even though there are relevant guidelines from the regulatory agencies, 52 they understandably do not direct the investigator to follow a specific methodology.…”

Lipid fraction of creams collected in the Parmigiano‐Reggiano cheese production area in response to extruded linseed supplementation of dairy cows’ diets: GC‐FID and FT‐MIR evaluation

“…To build a good ANN, it is important to determine a suitable number of neurons in the hidden layer. Generally, the number of neurons in the hidden layer is empirically determined depending on data quantity and behaviour; too many neurons can show a small training error but high generalisation error due to overfitting (Romani et al 2012). Therefore, different node numbers in the hidden layer momentum were tested, looking for the best classification ability.…”

Farm to Consumer: Factors Affecting the Organoleptic Characteristics of Coffee. II: Postharvest Processing Factors

“…An electronic nose is composed of an array of sensors which produce an array of signals (conductivity or resistance) to identify/quantify specific classes of volatiles for which they are made sensitive (Brattoli et al., 2003). Many studies have successfully applied e‐nose technology for distinguishing and classifying ground and instant coffee, commercial coffee blends, coffee mixtures, varieties, pure varieties, and espresso coffee aroma from different origins and roasting degrees (Romani, Cevoli, & Fabbri, ). Romani and coworkers also compared e‐nose technology with an artificial neural network for the prediction of roast degree, on the basis of the physical characteristics of roast degree.…”

“…The success rate of using electronic noses with cross‐validation was found to be 80% to 89.9%, with a need to lessen the number of sensor arrays (Falasconi, Pardo, & Sberveglieri, ). Later, the correlation of signals from an electronic nose with a reduced array of sensors for roast loss, L * a * b * , moisture, and density was set with the help of complex algorithms on large datasets (Romani et al., ). But still, the robust and reliable applicability of electronic noses has been found to be limited due to particular features like a large number of (descriptors) arrays (with respect to samples) having the same production technology, which can create the chance of correlations of fitting, leading to a poor predictability model (Roberts & Cozzolino ).…”