Prediction of problematic wine fermentations using artificial neural networks

Abstract

Artificial neural networks (ANNs) have been used for the recognition of non-linear patterns, a characteristic of bioprocesses like wine production. In this work, ANNs were tested to predict problems of wine fermentation. A database of about 20,000 data from industrial fermentations of Cabernet Sauvignon and 33 variables was used. Two different ways of inputting data into the model were studied, by points and by fermentation. Additionally, different sub-cases were studied by varying the predictor variables (total sugar, alcohol, glycerol, density, organic acids and nitrogen compounds) and the time of fermentation (72, 96 and 256 h). The input of data by fermentations gave better results than the input of data by points. In fact, it was possible to predict 100% of normal and problematic fermentations using three predictor variables: sugars, density and alcohol at 72 h (3 days). Overall, ANNs were capable of obtaining 80% of prediction using only one predictor variable at 72 h; however, it is recommended to add more fermentations to confirm this promising result.

Early recognition of problematic wine fermentations through multivariate data analyses

“…These results also agree closely with those reported by Román et al (2011) for application of artificial neural networks (ANNs) to the same data set, suggesting that both methods (MPLS and ANN) can perform equally well on the same set of data in this type of application. The previous must be compared and evaluated with an external validation with a different set of fermentations in order to establish if there are differences between the methods.…”

“…More recently, Urtubia, Perez-Correa, Soto, and Pszczokowski (2007) and Román, Hernandez, and Urtubia (2011) applied data mining and artificial neural network (ANN) techniques, respectively, to a large data set of variables from 24 batch fermentations of Cabernet Sauvignon for the purpose of predicting problematic batches early in the fermentation process. The work reported here makes use of this same data set by applying multiway principal component analysis (MPCA) and multiway partial least squares (MPLS) to unfold the set of multiway fermentation data, and compare the performance of one against the other, as well as with ANN, for early recognition of problematic wine fermentations.…”

Fault diagnosis of a benchmark fermentation process: a comparative study of feature extraction and classification techniques

“…Moreover, it is the opinion of the authors that for most commercial processes, suitable techniques exist for monitoring and diagnosis. As such, this work is a collaborative effort in view of establishing guidelines for the choice between some techniques, rather than extending the existing, which have been reported to perform successfully in bioprocesses [18,19]. This paper reports the first study in this line of research and the specific purpose is to find out whether the choice for MPCA or multi-way independent component analysis (MICA) as a feature extraction method and the choice for artificial neural networks (ANNs) or support vector machines (SVM) as a classification technique is important for the purpose of finding a suitable fault diagnosis strategy in batch processes.…”

Application of artificial neural networks coupled to UV–VIS–NIR spectroscopy for the rapid quantification of wine compounds in aqueous mixtures

“…ANNs have been used in many food technology applications, such as to predict the fatty acid composition of vegetable oils (Yalcin, Toker, Ozturk, Dogan, & Kisi, 2012), separate butters from different feedings and production seasons (Gori, Cevoli, Fabbri, Caboni, & Losi, 2012), determine the quality and freshness of pork (Huang, Zhao, Zhang, & Chen, 2012), determine the protein content or water absorption of cereals or flour (Mutlu et al, 2011), predict the antioxidant activity of tea (Cimpoiu, Cristea, Hosu, Sandru, & Seserman, 2011) or classify and predict beverages such as wines (Kruzlicova et al, 2009;Román et al, 2011), apple beverages (Gestal et al, 2004) or rice wines (Wei, Wang, & Ye, 2011).…”

“…The geographic origin is related to the wine quality, which is determined via sensorial analysis by oenologists and tasters or through expensive and time-consuming analytical methods. Furthermore, oenologists require more tools to improve the winemaking process and determine problems that can occur at all stages of winemaking (Román, Hernández, & Urtubia, 2011).…”