Classification of dry-cured hams according to the maturation time using near infrared spectra and artificial neural networks

An attempt to classify dry-cured hams according to the maturation time on the basis of near infrared (NIR) spectra was studied. The study comprised 128 samples of biceps femoris (BF) muscle from dry-cured hams matured for 10 (n = 32), 12 (n = 32), 14 (n = 32) or 16 months (n = 32). Samples were minced and scanned in the wavelength range from 400 to 2500 nm using spectrometer NIR System model 6500 (Silver Spring, MD, USA). Spectral data were used for i) splitting of samples into the training and test set using 2D Kohonen artificial neural networks (ANN) and for ii) construction of classification models using counter-propagation ANN (CP-ANN). Different models were tested, and the one selected was based on the lowest percentage of misclassified test samples (external validation). Overall correctness of the classification was 79.7%, which demonstrates practical relevance of using NIR spectroscopy and ANN for dry-cured ham processing control.     

Easy classification of traditional Minas cheeses using artificial neural networks and discriminant analysis

The classification of traditional Minas cheese (TMC) from different regions is important to ensure authenticity. Different chemometric approaches were used to discriminate TMCs from three different regions (Serro, Canastra and Araxá) of Minas Gerais, Brazil. The data obtained from the literature were used to develop an artificial neural network and to obtain linear discriminant functions, which were able to classify 100% of cheeses from different regions as a function of physico‐chemical composition. Both chemometric methods can be very useful tools to discriminate TMC from different regions based on physico‐chemical data which are obtained in a very quick and simple way.     

Antioxidant activity prediction and classification of some teas using artificial neural networks

In order to characterise and to classify some teas a simple, rapid and economical method based on composition, antioxidant activity and artificial neural networks (ANNs) is proposed. For these purpose two types of ANN based applications have been developed: one for predicting the antioxidant activity and a second one for establishing the class of the teas. The complex relationship between the total antioxidant activity (AA) depending on the total flavonoids content (F), total catechins content (C) and total methyl-xanthines content (MX) of commercial teas was revealed by the first designed feed-forward ANN. Secondly, using a probabilistic ANN, successful tea classification in various classes (green tea, black tea and express black tea) was also performed.     

A new approach for prediction of sewing performance of fabrics in apparel manufacturing using artificial neural networks

Abstract

This paper investigates the use of extended normalized radial basis function (ENRBF) neural networks to predict the sewing performance of fabrics in apparel manufacturing. In order to evaluate the performance of the ENRBF neural networks that could be emulated as human decision in the prediction of sewing performance of fabrics more effectively, it could be compared with the traditional back-propagation (BP) neural networks in terms of prediction errors. There are 109 data sets cover fabric properties measured by using a computerized measuring system, and the sewing performance of each fabric's specimen assessed by the domain experts. Of these 109 input—output data pairs, 94 were used to train the proposed ENRBF and BP neural networks for the prediction of the unknown sewing performance of a given fabric, and 15 were used to test the proposed ENRBF and BP neural networks, respectively. After 10,000 iterations of training of the ENRBF and BP neural networks, both of them converged to the minimum error level. A comparison was made between actual fabric performances during sewing, the experts' advices, and the results of predicting fabric performances during sewing for both networks. It was found that the ENRBF and BP neural networks indicate similar error levels, but the prediction made by the ENRBF neural network is better than the prediction made by the BP neural network in some areas. Both the systems provided better advice than the experts in some areas, when compared to actual sewing performance.     

A critical review on the use of artificial neural networks in olive oil production,characterization and authentication

Artificial neural networks (ANN) are computationally based mathematical tools inspired by the fundamental cell of the nervous system, the neuron. ANN constitute a simplified artificial replica of the human brain consisting of parallel processing neural elements similar to neurons in living beings. ANN is able to store large amounts of experimental information to be used for generalization with the aid of an appropriate prediction model. ANN has proved useful for a variety of biological, medical, economic and meteorological purposes, and in agro-food science and technology. The olive oil industry has a substantial weight in Mediterranean's economy. The different steps of the olive oil production process, which include olive tree and fruit care, fruit harvest, mechanical and chemical processing, and oil packaging have been examined in depth with a view to their optimization, and so have the authenticity, sensory properties and other quality-related properties of olive oil. This paper reviews existing literature on the use of bioinformatics predictive methods based on ANN in connection with the production, processing and characterization of olive oil. It examines the state of the art in bioinformatics tools for optimizing or predicting its quality with a view to identifying potential deficiencies or aspects for improvement.     

Characterisation of white vinegars of different sources with artificial neural networks

Wine and cider vinegars currently attract growing interest from consumers, giving rise to a consequent increase in supply. A full appreciation of their quality is only possible, however, through recognition of their superior quality when compared with vinegars produced from raw materials of inferior quality. Therefore, it is necessary to pinpoint the parameters that define the quality of these products. Chemico-physical and sensory analysis has been used to draw up artificial neural networks (ANNs), on the basis of a vast sampling of vinegars from various countries, produced from a variety of raw materials, that was already subjected to multivariate statistical analysis. Among the chemical parameters, polyalcohols and other elements such as pH, tartaric acid and proline proved to be highly reliable, whereas other volatile substances and the results of sensory analysis were not very discriminating and could not be used to re-classify samples of unknown origin. The positive results obtained indicate that ANNs are a powerful mathematical tool, since they can be used to construct models that predict the botanical origin of the product and to re-classify samples of unknown origin, without any initial restrictive hypothesis. © 1998 Society of Chemical Industry.     

Prediction of texture characteristics from extrusion food surface images using a computer vision system and artificial neural networks

Surface images and the texture characteristics of 17 samples and the 25 different parts within one sample were detected using a computer vision system and texture profile analysis in extruded food. According to the linear fitting model, the hardness and gumminess score can be reflected directly by the a^* and Intensity based on correlation coefficient of 0.9558, 0.9741 and 0.9429, 0.9619, respectively. The springiness could be reflected from color values through calculating from hardness and gumminess scores, indirectly. Neither of cohesiveness and chewiness presented relationship with two different color spaces. A desirable and accurate two hidden layers of back-propagation artificial neural network was trained for simulating and predicting the hardness and gumminess scores from a^* and Intensity based on the data in 17 samples, respectively. The simulation processing in ANN showed higher correlation coefficient of 0.9671 and 0.9856 than linear fitting model.     

An investigation on yarn engineering using artificial neural networks

Abstract

Engineering of spun yarns having specific tensile, evenness and hairiness characteristics is a long-cherished dream of spinning technologists. Selection of suitable raw materials at minimum cost and optimisation of process parameters are the two major tasks to be achieved to manufacture engineered yarn. Advent of high-speed fibre-testing machines and development of powerful modelling tools such as artificial neural network (ANN) have provided a great impetus in the yarn engineering research. This article demonstrates the feasibility of yarn engineering by developing a yarn-to-fibre ‘reverse’ model, using ANN. This approach is entirely different from the prevailing forward models, which predict the properties of final yarn using the fibre properties as inputs. The cost minimisation of cotton fibre mix was ensured by using the classical linear programming approach in combination with ANN. The engineered yarns demonstrated good agreement with the target yarn properties.     

Modeling the tryptic hydrolysis of pea proteins using an artificial neural network

Experimentally determined values for the degree of hydrolysis (DH) were used with an artificial neural network (ANN) model to predict the tryptic hydrolysis of a commercially available pea protein isolate at temperatures of 40, 45, and 50 °C. Analyses were conducted using the STATISTICA Neural Networks software on a personal computer. Input data were randomized to two sets: learning and testing. Differences between the experimental and calculated DH% were slight and ranged from 0.06% to 0.24%. The performance of the educated ANN was then tested by inputting temperatures ranging from 35 to 50 °C. Very strong correlations were found between calculated DH% values obtained from the ANN and those experimentally determined at all temperatures; the determination coefficients (R²) varied from 0.9958 to 0.9997. The results so obtained will be useful to reduce the time required in the design of enzymatic reactions involving food proteins.     

Simultaneous spectrophotometric determination of caffeine and theobromine in Iranian tea by artificial neural networks and its comparison with PLS

In order to determine the amount of caffeine and theobromine, spectrophotometry was used as a simple, rapid and economical method. Because of severe overlapping between these components, artificial neural network was used. The 230–300 nm spectral window with 1 nm interval was used for data acquisition. An artificial neural network (5-5-3) with linear transfer function between input-hidden and hidden-output layers was trained and applied for prediction of concentration of these methylxanthines in four Iranian tea samples. The model was compared with PLS modeling method. HPLC technique was used as a standard method.     

Quality evaluation of chickpeas using an artificial neural network integrated computer vision system

Chickpea is one of the most consumed legumes in the world. The classification of chickpea based on the size and morphological properties is important for the market. The objective of this study is to design and implement a computer vision system (CVS) integrated with artificial neural networks (ANN) for quality evaluation of chickpeas based on their size, colour, and surface morphology. The system is composed of a flat bed scanner for acquiring digital image and software that has been developed in Matlab for image analysis. Physical properties (length, width and volume) of the samples of chickpeas as well as their colour properties and surface characteristics have been determined by using the system, and results have been validated. High correlations have been found between the results from ANN‐integrated CVS and those obtained by callipers or professionally trained inspectors based on the experiments. Overall, percentages of correct classification have been determined as 95.4%, 87.6%, and 96.0% for colour, surface morphology, and shape evaluations, respectively.     

An attempt to predict pork drip loss from pH and colour measurements or near infrared spectra using artificial neural networks

The ability to predict meat drip loss by using either near infrared spectra (SPECTRA) or different meat quality (MQ) measurements, such as pH₂₄, Minolta L^∗, a^∗, b^∗, along with different chemometric approach, was investigated. Back propagation (BP) and counter propagation (CP) artificial neural networks (ANN) were used and compared to PLS (partial least squares) regression. Prediction models were created either by using MQ measurements or by using NIR spectral data as independent predictive variables. The analysis consisted of 312 samples of longissimus dorsi muscle. Data were split into training and test set using 2D Kohonen map. The error of drip loss prediction was similar for ANN (2.2–2.6%) and PLS models (2.2–2.5%) and it was higher for SPECTRA (2.5–2.6%) than for MQ (2.2–2.3%) based models. Nevertheless, the SPECTRA based models gave reasonable prediction errors and due to their simplicity of data acquisition represent an acceptable alternative to classical meat quality based models.     

Aroma components as discriminating parameters in the chemometric classification of venetian white wines

Twenty-two volatile aroma components of three Venetian white wines (vintage of 1981) were determined by capillary gas chromatography, after liquid-liquid extraction and concentration in a Vigreux column. Multiple discriminant analysis and the k-nearest neighbour rule were applied to data in order to differentiate and classify the wines according to origin. Using 11 selected parameters (alcoholic and acidic fractions), the correctly classified cases were 94.9% by the multiple discriminant analysis, and 71.8% by the 3-nearest neighbour rule.     

Estimation of somatic cell count levels of hard cheeses using physicochemical composition and artificial neural networks

This study addresses the prediction of the somatic cell counts of the milk used in the production of sheep cheese using artificial neural networks. To achieve this objective, the neural network was designed using 33 parameters of the physicochemical composition of the cheeses obtained after they have been matured for 12 mo as input data. The physicochemical analysis of the cheeses revealed that the somatic cell count level of the cheese has a significant influence on the amount of protein, fat, dry extract, and fatty acids. When properly set up, the neural network allows the correct classification of the cheeses (100% of correct results in both training and test phases) and therefore their samples in each of the 3 nominal output variables (low, average, and high somatic cell counts). The fatty composition of the cheeses, individual fatty acids, and fat acidity are the variables that most affect the correct operation of the neural network.     

Verification of the geological origin of bottled mineral water using artificial neural networks

As a first step towards objective and cost-efficient verification of the geographical origin of commercially sold mineral water, we determined up to what extent the chemical composition of mineral water can be linked to the geology of the local water source. For this purpose, a dataset consisting of 145 European mineral water samples from a known geology was analysed using counter-propagation artificial neural networks (CP-ANNs) with supervised learning algorithm. The models were tested for recall ability (RA) and validated with a leave-one-out cross validation (LOO-CV).     

Prediction of colloidal stability in white wines using infrared spectroscopy

The feasibility of utilizing infrared spectroscopy for the prediction of haze formation in white wines resulting from heat and colloidal stability tests was investigated. One-hundred eleven white wines, representing multiple regions and varieties from the 2008 California vintage, were collected and analyzed. The near and mid-infrared spectra were measured and heat and colloidal (ethanol addition) stability tests were performed on the same wines. Partial-least squares regression analysis was then used to construct models predictive of the resulting nepholometric turbidity to the acquired spectra. Preliminary models obtained following application of spectral pretreatments today considered as “classical” (e.g., derivatives, standard normal variate, vector normalization, constant offset elimination) lacked robustness; two alternative algorithms designed to remove spectral information unrelated to the turbidity were then employed (orthogonal signal correction; direct orthogonal signal correction). While OSC pretreatment did not result in more robust models, DOSC considerably enhanced the goodness of the PLS model constructed to predict the ethanol test turbidity. Predictive modeling of the short-NIR spectra, following DOSC preprocessing, allowed the prediction of colloidal stability on an unknown test set with an R² = 0.80 and a RMSEP = 10.12 using three latent variables. When the data set was restricted to Chardonnay wines alone, the predictive ability improved, with R² = 0.85 and RMSEP = 8.90.     

BP神经网络在纺织工业中的应用

介绍了神经网络的发展历史,对BP网络的算法进行了详细的讨论,分析了BP网络在纺织工业中应用的主要步骤。     

利用AFIS与神经网络预测纱线强力

评述了AFIS纤维测试系统以及人工神经网络的特点，提出了利用它们进行纱线强力预测的工作原理和网络构建方法，并提供了实例。     

基于人工神经网络的毛精纺纱线质量预报技术

介绍了毛精纺纺纱过程与人工神经网络的特点 ,提出人工神经网络在纺纱质量预报中的工作原理与实现方法 ,并提供了国内外的应用实例 ,指出人工神经网络技术在毛精纺纱线质量预报中的广泛应用前景。