Optimization of an artificial neural network for thermal/pressure food processing: Evaluation of training algorithms

The aim of the current paper is to obtain, through a proper selection of the training algorithm, an optimized artificial neural network (ANN) able to predict two parameters of interest for high-pressure (HP) food processing: the maximum or minimum temperature reached in the sample after pressurization and the time needed for thermal re-equilibration in the high-pressure process. To do that, 13 training algorithms belonging to 4 broad classes (gradient descent, conjugate gradient, quasi-Newton algorithms and Bayesian regularization) have been evaluated by training different ANNs. The network trained with the Levenberg-Marquardt algorithm showed the best overall predictive ability. The performance of this network was subsequently optimized by varying the number of nodes in the hidden layer, the learning coefficient and the decrease factor of this coefficient, and selecting the configuration with the highest predictive ability. The optimized ANN was able to make accurate predictions for the variables studied (temperature and time). These predictions were significantly better than those obtained by a previous ANN developed without selection of the training algorithm, that is, assuming the default option of the ANN computational package (gradient descent with a user-defined learning rate). We have shown that a correct selection of the training algorithm allows maximizing the predictive ability of the artificial neural network.

[1]  José S. Torrecilla,et al.  A neural network approach for thermal/pressure food processing , 2004 .

[2]  Xiukun Yang,et al.  Use of genetic artificial neural networks and spectral imaging for defect detection on cherries , 2000 .

[3]  Laura Otero,et al.  A model to design high-pressure processes towards an uniform temperature distribution , 2007 .

[4]  José S. Torrecilla,et al.  Artificial neural networks : a promising tool to design and optimize high-pressure food processes , 2005 .

[5]  M. C. Palancar,et al.  pH-Control System Based on Artificial Neural Networks , 1998 .

[6]  Laura Otero,et al.  Modelling heat transfer in high pressure food processing: a review , 2003 .

[7]  Laura Otero,et al.  Thermal effect in foods during quasi-adiabatic pressure treatments , 2000 .

[8]  Ann Van Loey,et al.  A modeling approach for evaluating process uniformity during batch high hydrostatic pressure processing: combination of a numerical heat transfer model and enzyme inactivation kinetics , 2000 .

[9]  Sundaram Gunasekaran,et al.  Food quality prediction with neural networks , 1998 .

[10]  V. M. Balasubramaniam,et al.  Determining Thermal Effects in High-Pressure Processing , 2002 .

[11]  Peter York,et al.  Optimisation of the predictive ability of artificial neural network (ANN) models: a comparison of three ANN programs and four classes of training algorithm. , 2005, European journal of pharmaceutical sciences : official journal of the European Federation for Pharmaceutical Sciences.

[12]  Laura Otero,et al.  A Model for Real Thermal Control in High‐Pressure Treatment of Foods , 2002, Biotechnology progress.

[13]  Roger Ruan,et al.  Prediction of Dough Rheological Properties Using Neural Networks , 1995 .

[14]  Dallas G. Hoover,et al.  High Pressure Processing , 2000 .

[15]  Lewis B. Sheiner,et al.  Some suggestions for measuring predictive performance , 1981, Journal of Pharmacokinetics and Biopharmaceutics.

[16]  Shiv O. Prasher,et al.  Thermal process calculations using artificial neural network models , 2001 .

[17]  A H Geeraerd,et al.  Application of artificial neural networks as a non-linear modular modeling technique to describe bacterial growth in chilled food products. , 1998, International journal of food microbiology.