DSpace Arşivi :: by Yazar "Tarlak, Fatih" değerine göre listeleniyor

Yazar "Tarlak, Fatih" seçeneğine göre listele

Listeleniyor 1 - 20 / 26

A new expanded modelling approach for investigating the bioprotective capacity of Latilactobacillus sakei CTC494 against Listeria monocytogenes in ready-to-eat fish products
(Elsevier, 2021) Bolívar, Araceli; Tarlak, Fatih; Correia Peres Costa, Jean Carlos; Cejudo-Gomez, Manuel; Bover-Cid, Sara; Zurera, Gonzalo; Perez-Rodriguez, Fernando
Understanding the role of food-related factors on the efficacy of protective cultures is essential to attain optimal results for developing biopreservation-based strategies. The aim of this work was to assess and model growth of Latilactobacillus sakei CTC494 and Listeria monocytogenes CTC1034, and their interaction, in two different ready-to-eat fish products (i.e., surimi-based product and tuna pate) at 2 and 12 degrees C. The existing expanded Jameson-effect and a new expanded Jameson-effect model proposed in this study were evaluated to quantitatively describe the effect of microbial interaction. The inhibiting effect of the selected lactic acid bacteria strain on the pathogen growth was product dependent. In surimi product, a reduction of lag time of both strains was observed when growing in coculture at 2 degrees C, followed by the inhibition of the pathogen when the bioprotective L. sakei CTC494 reached the maximum population density, suggesting a mutualism-antagonism continuum phenomenon between populations. In tuna pate, L. sakei CTC494 exerted a strong inhibition of L. monocytogenes at 2 degrees C (<0.5 log increase) and limited the growth at 12 degrees C (<2 log increase). The goodness-of-fit indexes indicated that the new expanded Jameson-effect model performed better and appropriately described the different competition patterns observed in the tested fish products. The proposed expanded competition model allowed for description of not only antagonistic but also mutualism-based interactions based on their influence on lag time.
An intelligent based prediction of microbial behaviour in beef
(Elsevier Sci Ltd, 2023) Yucel, Ozgun; Tarlak, Fatih
The main purpose of this work was to develop different machine learning-based regression methods referred to as decision tree regression (DTR), generalized additive model regression (GAMR) and random forest regression (RFR) to predict bacterial population on beef. For this purpose, 2654 bacterial data points of Listeria monocytogenes, Escherichia coli and Pseudomonas spp. Which are the most investigated bacterial genera in beef were collected from the ComBase database (www.combase.cc). Temperature, salt concentration, water activity and acidity were used as the main predictor variables to estimate the growth or survival behaviour of the microorganisms in beef. The hyperparameters are optimized for proposed machine learning-based regression methods with nested cross-validation. The fitting capabilities of the proposed machine learning algorithms were compared considering their statistical indices (coefficient of determination R2 and root mean square error RMSE). Each regression method provided satisfactory predictions with being 0.931 < R2 < 0.949 and 0.597 < RMSE < 0.692 considering each of the microorganism populations. However, the RFR yielded the best prediction capability and therefore its prediction capability was further assessed. The RFR in the external validation process provided statistical indices being 1.017 < Bias factor <1.151 and 1.137 < Accuracy factor <1.370, indicating that the random forest regression can be reliably employed as an alternative way of describing simultaneously survival and growth behaviour of microorganisms in beef and has a significant potential to be used as an alternative simulation method by skipping a secondary model step in two-step modelling approach, traditionally utilized in the predictive microbiology field.
Application of a machine learning-based regression method to describe Listeria monocytogenes behaviour in milk
(Vup Food Research Inst, Bratislava, 2022) Tarlak, Fatih; Yucel, Ozgun
The main aim of the present study was to develop a prediction tool to describe Listeria monocytogenes behaviour in milk by employing traditionally used models (the re-parametrized Gompertz, Baranyi and Huang models) and an alterna-tively proposed machine learning-based regression model. The fitting capability of both groups of models was evaluated and compared considering their statistical indices (coefficient of determination R-2, root mean square error RMSE). The machine learning-based regression model provided better predictions (with R-2 of 0.958 and RMSE of 0.407) than the traditionally used models. The prediction capability of both methodologies was tested considering externally collected data from the literature. The machine learning-based regression model in the validation process gave satisfactory sta-tistical indices (bias factor of 1.016 and accuracy factor of 1.056), which is better prediction power than the traditionally used models. These results indicated that the machine learning-based regression method can be reliably employed as an alternative way of describing the growth behaviour of L. monocytogenes in milk. Therefore, the software developed in this work has a significant potential to be used as an alternative simulation method to traditionally used approach in the field of predictive microbiology.
Comparison of modelling approaches for the prediction of kinetic growth parameters of Pseudomonas spp. in oyster mushroom (Pleurotus ostreatus)
(Sage Publications Ltd, 2023) Tarlak, Fatih; Correia Peres Costa, Jean Carlos
In predictive microbiology, primary and secondary models can be used to predict microbial growth, usually in a two-step modelling approach. The inverse dynamic modelling approach is an alternative method to direct modelling methods, in which the primary and secondary models are fitted simultaneously from non-isothermal data, minimising experimental effort and costs. Thus, the main aim of the present study was to compare the prediction capabilities of the mathematical modelling approaches used for calculating growth kinetics of microorganisms in predictive food microbiology field. For this purpose, the bacterial growth data of Pseudomonas spp. in oyster mushroom (Pleurotus ostreatus) subjected to isothermal and non-isothermal storage temperatures were collected from previously published growth curves. Temperature-dependent kinetic growth parameters (maximum specific growth rate 'mu (max) ' and lag phase duration 'lambda') were described as a function of storage temperature using the direct two-step, direct one-step and inverse dynamic modelling approach based on Baranyi and Huang models. The fitting capability of the modelling approaches was separately compared, and the one-step modelling approach for the direct methods provided better goodness of fit results regardless of used primary models, which leads the Huang model with being RMSE = 0.226 and R-adj(2) = 0.949 became best for direct methods. Like seen in direct methods, the Huang model gave better goodness of fit results than Baranyi model for inverse method. Results revealed there was no significant difference (p > 0.05) between the growth kinetic parameters obtained from direct one-step modelling approach and inverse modelling approaches based on the Huang model. Satisfactorily statistical indexes show that the inverse dynamic modelling approach can be reliably used as an alternative way of describing the growth behaviour of Pseudomonas spp. in oyster mushroom in a fast and minimum labour effort.
Development and validation of a one-step modelling approach for the determination of chicken meat shelf-life based on the growth kinetics of Pseudomonas spp
(Sage Publications Ltd, 2022) Tarlak, Fatih; Perez-Rodriguez, Fernando
The main objective of the present study was to investigate the effect of storage temperature on aerobically stored chicken meat spoilage using the two-step and one-step modelling approaches involving different primary models namely the modified Gompertz, logistic, Baranyi and Huang models. For this purpose, growth data points of Pseudomonas spp. were collected from published studies conducted in aerobically stored chicken meat product. Temperature-dependent kinetic parameters (maximum specific growth rate 'mu (max) ' and lag phase duration 'lambda') were described as a function of storage temperature through the Ratkowsky model based on the different primary models. Then, the fitting capability of both modelling approaches was compared taking into account root mean square error, adjusted coefficient of determination (adjusted-R-2) and corrected Akaike information criterion. The one-step modelling approach showed considerably improved fitting capability regardless of the used primary model. Finally, models developed from the one-step modelling approach were validated for the maximum growth rate data extracted from independent published literature using the statistical indexes Bias (B-f) and Accuracy (A(f)) factors. The best prediction capability was obtained for the Baranyi model with B-f and A(f) being very close to 1. The shelf-life of chicken meat as a function of storage temperature was predicted using both modelling approaches for the Baranyi model.
Development and validation of one-step modelling approach for prediction of mushroom spoilage
(Vup Food Research Inst, Bratislava, 2020) Tarlak, Fatih
The primary aims of this work were to improve the prediction capability of the traditionally used two-step modelling approach with the most popular primary growth models in the predictive food microbiology field, and to validate the prediction capability of the one-step modelling approach, a proposed alternative way to traditional modelling approach. For this purpose, the growth behaviour of Pseudomonas spp. existing in the natural microflora of button mushrooms (Agaricus bisporus) was simulated with two-step and one-step modelling approaches. The Baranyi model yielded the best fitting performance when it was employed in the two-step modelling approach. The fitting capability of all the primary models was also compared using the one-step modelling approach. No matter which primary model was used, the one-step modelling approach significantly improved the prediction capability of the models, and all the primary models gave root mean squared error lower than 0.299 and adjusted coefficient of determination higher than 0.948. Successfully validated Baranyi model in one-step modelling approach provided the highest prediction capability and exhibited considerable potential to be used as a prediction tool. This indicated that the one-step modelling approach could be reliably employed to assess and predict mushroom spoilage as a function of time and storage temperature.
Development of a new mathematical modelling approach for prediction of growth kinetics of Listeria monocytogenes in milk
(Vup Food Research Inst, Bratislava, 2021) Tarlak, Fatih
The main objective of the present study was to develop a new modelling method, inverse dynamic modelling approach, as an alternative to two-step modelling approach, which is traditionally used n predictive food microbiology. For this purpose, the growth data of Listeria monocytogenes in milk subjected to isothermal and non-isothermal storage conditions were gathered from previously published growth curves. The bacterial growth data were described as a function of time and temperature using the direct two-step, direct one-step and inverse dynamic modelling approaches based on the Baranyi and Huang models. Maximum specific growth rate (mu(max)) and lag phase duration (lambda) estimated by different modelling approaches and primary models were statistically compared. Results revealed that there was no significant difference (p > 0.05) between the growth kinetic parameters obtained from direct and inverse modelling approaches. The prediction capability of inverse dynamic modelling approach was validated by externally gathering growth curves. The inverse dynamic modelling approach provided satisfactory statistical indices (0.99 > Bias factor > 1.10 and 1.16 > Accuracy factor > 1.19), meaning that it can be reliably used as an alternative way of describing the growth behaviour of Listeria monocytogenes in milk in a fast way with a minimal labour requirement.
Development of a New Modelling Approach and Performance Evaluation of Meta-heuristic Optimization Algorithms for the Prediction of Kinetic Growth Parameters for Pseudomonas spp. in Fish
(Dr M N Khan, 2022) Tarlak, Fatih; Yucel, Ozgun; Khosravi-Darani, Kianoush
The main aim of the current work was to build up a new mathematical modelling approach in predictive food microbiology field for the prediction of growth kinetics of microorganisms. For this purpose, the bacterial growth data of Pseudomonas spp. in whole fish (gilt-head seabream) subjected to isothermal and non-isothermal storage temperatures were collected from previously published growth curves. Maximum specific growth rate (1/h) and lag phase duration (h) were described as a function of storage temperature using the direct two-step, direct one-step and inverse dynamic modelling approaches based on various meta-heuristic optimization algorithms. The fitting capability of the modelling approaches and employed optimization algorithms was separately compared, and the one-step modelling approach for the direct methods and the Bayesian optimization method for the used algorithms provided the best goodness of fit results. These two were then further processed in validation step. The inverse dynamic modelling approach based on the Bayesian optimization algorithm yielded satisfactorily statistical indexes (1.02 > Bias factor > 1.09 and 1.07 > Accuracy factor > 1.13), which indicates it can be reliably used as an alternative way of describing the growth behaviour of Pseudomonas spp. in fish in a fast and efficient manner with minimum labour effort.
Development of a Novel Growth Model Based on the Central Limit Theorem for the Determination of Beef Spoilage
(Shahid Beheshti Univ Medical Sciences, Fac Med, 2021) Tarlak, Fatih
Background and Objective: Currently, no published studies are available that compare central limit theorem model with traditionally used growth models in predictive food microbiology to describe bacterial growth behaviors of Pseudomonas spp. in beefs. The major objectives of the present study were to develop a novel growth model based on the central limit theorem and compare the prediction capability of the model with those of various growth models (modified Gompertz, logistic, Baranyi and Huang models) commonly used in predictive food microbiology. Material and Methods: Bacterial growth data for Pseudomonas spp. were collected from previously published studies on beefs stored at isothermal storage temperatures (0, 4, 7, 10, 15 and 20 degrees C). Temperature dependent kinetic parameters (maximum specific growth rate 'mu(max)' and lag phase duration 'lambda') collected from various primary models were described as functions of storage temperatures using Ratkowsky model. Fitting capability of the novel growth model based on the central limit theorem was compared with other growth models using mean square error and coefficient of determination. Results and Conclusion: The novel growth model developed in this study provided mean square errors less than 0.104 and coefficients of determination greater than 0.962. No significant differences (p>0.05) were seen between the statistical indices of this developed model and traditionally used growth models. Results have shown that the novel growth model based on the central limit theorem can be used to describe the growth behaviors of microorganisms as alternative to traditionally used growth models of modified Gompertz, logistic, Baranyi and Huang models in predictive food microbiology. Furthermore, this novel model can be used for the prediction of shelf-life of beefs as a function of temperature since spoilage of beefs is directly linked to the load of Pseudomonas spp.
Development of a Prediction Software for the Growth Kinetics of Pseudomonas spp. in Culture Media using Various Primary Models
(Shahid Beheshti Univ Medical Sciences, Fac Med, 2023) Tarlak, Fatih; Yucel, Ozgun
Background and Objective: Pseudomonas spp. are bacteria with the widest effects on food spoilage. These bacteria can be found in several environments such as soil and water. The major purpose of this study was to develop a software; by which, the growth behaviours of Pseudomonas spp. in culture media could be predicted.Material and Methods: A total number of 509 bacterial data points of Pseudomonas spp. in culture media were collected from the ComBase database. Temperature and pH were used as the major prediction variables for the description of Pseudomonas spp. behaviours in culture media. Modified Gompertz, Baranyi and Huang models, the most commonly used models in predictive food microbiology to predict the count of microorganisms, were used as well. Fitting capability of each model was assessed and compared with other capabilities considering their statistical indices of the root mean square error, RMSE; coefficient of determination, R2; corrected Akaike information criterion, AICc; and Bayesian information criterion, BIC.Results and Conclusion: Huang model provided better predictions with 0.951 of R2 and 0.825 of RMSE, compared to those of traditionally used models. Prediction capability of the Huang model was assessed considering externally collected data from the ComBase database. Huang model in the validation process provided satisfactory statistical indices (bias factor = 1.027 and accuracy factor = 1.075). These results have revealed that Huang model can be reliably used as a model of describing the growth behaviours of Pseudomonas spp. Furthermore, developed software in this study includes significant potentials for predicting Pseudomonas counts in culture media.
Development of prediction software to describe total mesophilic bacteria in spinach using a machine learning-based regression approach
(Sage Publications Ltd, 2023) Yildirim-Yalcin, Meral; Yucel, Ozgun; Tarlak, Fatih
The purpose of this study was to create a tool for predicting the growth of total mesophilic bacteria in spinach using machine learning-based regression models such as support vector regression, decision tree regression, and Gaussian process regression. The performance of these models was compared to traditionally used models (modified Gompertz, Baranyi, and Huang models) using statistical indices like the coefficient of determination (R-2) and root mean square error (RMSE). The results showed that the machine learning-based regression models provided more accurate predictions with an R-2 of at least 0.960 and an RMSE of at most 0.154, indicating that they can be used as an alternative to traditional approaches for predictive total mesophilic. Therefore, the developed software in this work has a significant potential to be used as an alternative simulation method to traditionally used approach in the predictive food microbiology field.
FOOD RECOGNITION AND NUTRITION FACTS DETERMINATION WITH DEEP CONVOLUTION NEURAL NETWORK MODELS
(North Univ Center Baia Mare, 2023) Tarlak, Fatih; Yucel, Ozgun; Yilmaz, Onur
Food recognition plays a crucial role in various domains including healthcare, nutrition, and the food industry. In healthcare, food recognition is valuable for individuals to monitor their daily food intake and manage their diet effectively. It also aids dietitians and nutritionists in creating personalized meal plans for patients based on their nutritional requirements and preferences. The objective of this research was to develop software capable of recognizing and predicting the nutritional information of commonly consumed fruits and vegetables in Turkey. Basic nutrition data for each food item was collected and organized. A dataset comprising 9,000 food images was gathered, encompassing 200 images for each food item. To train the images, deep learning (DL) algorithms such as GoogleNet, ResNet-50, and Inception-v3 were utilized on platforms like Matlab and.NET Core. Additionally, 900 food images were reserved for external validation purposes. The DL algorithms achieved excellent accuracy, with all models surpassing 98.3% accuracy in predicting food categories. Notably, the Inception-v3 algorithm outperformed the others, achieving an accuracy of 99.1% during the testing phase. Consequently, the Inception-v3 algorithm was chosen to develop software for food recognition and nutrition analysis, intended for both computers and smartphones. The software can be relied upon for food recognition and nutrition analysis, making it highly valuable in healthcare, particularly in tracking the dietary intake of patients with chronic conditions like diabetes, heart disease, or obesity. The system can effectively track the types and quantities of foods consumed, providing personalized feedback to both patients and healthcare providers.
Inactivation of Salmonella Typhimurium in fresh-cut lettuce during chlorine washing: Assessing the impacts of free chlorine concentrations and exposure times
(Elsevier, 2023) Possas, Aricia; Posada-Izquierdo, Guiomar Denisse; Tarlak, Fatih; Jimenez-Jimenez, Francisco; Perez-Rodriguez, Fernando
The aim of this study was to evaluate, quantify and model the inactivation of Salmonella in fresh-cut lettuce during washings with chlorinated water at different free chlorine concentrations (FCC, 0-150 mg/L). Individual fresh-cut lettuce samples (4 cm2) were inoculated with a Salmonella culture (ca. 4 log CFU/cm2) and washed with 100-mL solutions with different FCC for different times (0-150 s). The surviving Salmonella cells recovered from samples were enumerated by plate count methodology. A fast decay on Salmonella counts was marked in the first 20 s of washing, followed by a slowing down on reductions. A maximum of 2.6 log-decrease was observed after 2.5-min washing regardless of FCC. The log-linear with tail primary model coupled with a linear secondary model was fitted to inactivation data obtained at FCC from 50 to 150 mg/L through global regression analysis, yielding a suitable model to describe Salmonella concentrations as a function of FCC and washing times (RMSE = 0.34, R2adj= 0.84). Simulations using the developed model showed inactivation rates varying from 0.17 log CFU/ s at 50 mg/L to 0.86 log CFU/s at 150 mg/L. Disinfection models on lettuce are valuable tools for the validation of control measures in the fresh-cut produce industry and for quantitative risk assessments.
MICROORGANISMS RESPONSIBLE FOR DETERIORATION OF FOOD PRODUCTS: REVIEW
(North Univ Center Baia Mare, 2022) Pinarli, Cagla; Tarlak, Fatih
Every year, tons of food is thrown away because of changes in odor, taste, texture or color. Food spoilage has a very important effect at this point. Microorganisms have a significant position in food spoilage. The type of microorganisms seen varies depending on factors such as the amount of water, acidity, carbohydrate, protein or fat ratios of foods, packaging type, and oxygen levels. In this section, basic organisms that cause food spoilage (bacteria, mold, yeast) and microorganisms that are effective in specific food groups (meat products, poultry products, dairy products, seafood products, egg products, cereal products, fruits and vegetables) are covered. Approaches to reduce spoilage organisms are as important as identifying spoilage organisms. Therefore, recommendations for controlling and preventing spoilage organisms are also included in this section.
Modeling the Growth of Six Listeria monocytogenes Strains in Smoked Salmon Pate
(Mdpi, 2023) Bolivar, Araceli; Garrote Achou, Chajira; Tarlak, Fatih; Cantalejo, Maria Jesus; Costa, Jean Carlos Correia Peres; Perez-Rodriguez, Fernando
In this study, the growth of six L. monocytogenes strains isolated from different fish products was quantified and modeled in smoked salmon pate at a temperature ranging from 2 to 20 degrees C. The experimental data obtained for each strain was fitted to the primary growth model of Baranyi and Roberts to estimate the following kinetic parameters: lag phase (lambda), maximum specific growth rate (mu(max)), and maximum cell density (N-max). Then, the effect of storage temperature on the obtained mu(max) values was modeled by the Ratkowsky secondary model. In general, the six L. monocytogenes strains showed rapid growth in salmon pate at all storage temperatures, with a relatively short lag phase lambda, even at 2 degrees C. The growth behavior among the tested strains was similar at the same storage temperature, although significant differences were found for the parameters lambda and mu(max). Besides, the growth variations among the strains did not follow a regular pattern. The estimated secondary model parameter T-min ranged from -4.25 to -3.19 degrees C. This study provides accurate predictive models for the growth of L. monocytogenes in fish pates that can be used in shelf life and microbial risk assessment studies. In addition, the models generated in this work can be implemented in predictive modeling tools and repositories that can be reliably and easily used by the fish industry and end-users to establish measures aimed at controlling the growth of L. monocytogenes in fish-based pates.
Modelling of the Behaviour of Salmonella enterica serovar Reading on Commercial Fresh-Cut Iceberg Lettuce Stored at Different Temperatures
(Mdpi, 2020) Tarlak, Fatih; Johannessen, Gro; Bascon Villegas, Isabel; Bolivar, Araceli; Posada-Izquierdo, Guiomar Denisse; Perez-Rodriguez, Fernando
The aim of this study was to model the growth and survival behaviour ofSalmonellaReading and endogenous lactic acid bacteria on fresh pre-cut iceberg lettuce stored under modified atmosphere packaging for 10 days at different temperatures (4, 8 and 15 degrees C). The Baranyi and Weibull models were satisfactorily fitted to describe microbial growth and survival behaviour, respectively. Results indicated that lactic acid bacteria (LAB) could grow at all storage temperatures, whileS. Reading grew only at 15 degrees C. Specific growth rate values (mu(max)) for LAB ranged between 0.080 and 0.168 h(-1)corresponding to the temperatures 4 and 15 degrees C while forS. Reading at 15 degrees C,mu(max)= 0.056 h(-1). This result was compared with published predictive microbiology models for otherSalmonellaserovars in leafy greens, revealing that predictions from specific models could be valid for such a temperature, provided they were developed specifically in lettuce regardless of the type of serovars inoculated. The parameter delta obtained from the Weibull model for the pathogen was found to be 16.03 and 18.81 for 4 and 8 degrees C, respectively, indicating that the pathogen underwent larger reduction levels at lower temperatures (2.8 log(10)decrease at 4 degrees C). These data suggest that thisSalmonellaserovar is especially sensitive to low temperatures, under the assayed conditions, while showcasing that a correct refrigeration could be an effective measure to control microbial risk in commercial packaged lettuce. Finally, the microbiological data and models from this study will be useful to consider more specifically the behaviour ofS. Reading during transport and storage of fresh-cut lettuce, elucidating the contribution of this serovar to the risk bySalmonellain leafy green products.
Prediction of growth kinetics of Pseudomonas spp. in meat products under isothermal and non-isothermal storage conditions
(2021) Tarlak, Fatih
The main objective of the present study was to develop and validate a new alternative modelling method to predict the shelf-life of food products under non-isothermal storage conditions. The bacterial growth data of the Pseudomonas spp. was extracted from published studies conducted for aerobically-stored fish, pork and chicken meat and described with two-step and one-step modelling approaches employing different primary models (the modified Gompertz, logistic, Baranyi and Huang models) under isothermal storage temperatures. Temperature dependent kinetic parameters (maximum specific growth rate ‘µmax’ and lag phase duration ‘?’) were described as a function of storage temperature via the Ratkowsky model integrated with each primary model. The Huang model based on the one-step modelling approach yielded the best goodness of fit results ($RMSE = 0.451 and adjusted-R^2 = 0.942$) for all food products at isothermal storage conditions, therefore, was also used to check it’s the prediction capability under non-isothermal storage conditions. The differential form of the Huang model provided satisfactorily statistical indexes ($1.075 > B_f > 1.014 and 1.080 > A_f > 1.047$) indicating reliably being able to use to describe the growth behaviour of Pseudomonas spp. in fish, pork and chicken meat subjected to non-isothermal storage conditions.
Prediction of indigenous Pseudomonas spp. growth on oyster mushrooms (Pleurotus ostreatus) as a function of storage temperature
(Elsevier, 2019) Manthou, Evanthia; Tarlak, Fatih; Lianou, Alexandra; Ozdemir, Murat; Zervakis, Georgios I.; Panagou, Efstathios Z.; Nychas, George-John E.
The growth kinetic behaviour of Pseudomonas spp. naturally occurring on oyster mushrooms (Pleurotus ostreatus) was evaluated during storage at different isothermal conditions (4, 10 and 16 degrees C), and was described quantitatively using a one-step global parameter estimation method. In the context of this modelling approach, the growth kinetic parameters of maximum specific growth rate (mu(max)) and lag phase duration (lambda) were estimated using the Baranyi model, whereas the effect of temperature mu(max) was described using a secondary square-root type model. The global model's goodness-of-fit indices of root mean square error (RMSE) and adjusted coefficient of determination (adjusted-R-2) were estimated to be 0.206 and 0.948, respectively. The global model was then externally validated using growth data generated during storage of oyster mushrooms under dynamic temperature conditions. Specifically, the differential form of the Baranyi model merged with the square-root-type model was solved numerically using the fourth-order Runge-Kutta method in order to predict the Pseudomonas spp. concentration on mushrooms under fluctuating temperature conditions. The developed dynamic modelling approach exhibited satisfactory performance, with the mean deviation and the mean absolute deviation being -0.10 and 0.22 log CFU/g, respectively. Along with further substantiation and optimization, the developed model should be useful in food quality management systems, aiming in particular at the improvement of the microbiological quality of oyster mushrooms.
Prediction of Pseudomonas spp. Population in Food Products and Culture Media Using Machine Learning-Based Regression Methods
(Mdpi, 2023) Tarlak, Fatih; Yucel, Ozgun
Machine learning approaches are alternative modelling techniques to traditional modelling equations used in predictive food microbiology and utilise algorithms to analyse large datasets that contain information about microbial growth or survival in various food matrices. These approaches leverage the power of algorithms to extract insights from the data and make predictions regarding the behaviour of microorganisms in different food environments. The objective of this study was to apply various machine learning-based regression methods, including support vector regression (SVR), Gaussian process regression (GPR), decision tree regression (DTR), and random forest regression (RFR), to estimate bacterial populations. In order to achieve this, a total of 5618 data points for Pseudomonas spp. present in food products (beef, pork, and poultry) and culture media were gathered from the ComBase database. The machine learning algorithms were applied to predict the growth or survival behaviour of Pseudomonas spp. in food products and culture media by considering predictor variables such as temperature, salt concentration, water activity, and acidity. The suitability of the algorithms was assessed using statistical measures such as coefficient of determination (R-2), root mean square error (RMSE), bias factor (Bf), and accuracy (A(f)). Each of the regression algorithms showed appropriate estimation capabilities with R-2 ranging from 0.886 to 0.913, RMSE from 0.724 to 0.899, B-f from 1.012 to 1.020, and A(f) from 1.086 to 1.101 for each food product and culture medium. Since the predictive capability of RFR was the best among the algorithms, externally collected data from the literature were used for RFR. The external validation process showed statistical indices of B-f ranging from 0.951 to 1.040 and A(f) ranging from 1.091 to 1.130, indicating that RFR can be used for predicting the survival and growth of microorganisms in food products. Therefore, machine learning approaches can be considered as an alternative to conventional modelling methods in predictive microbiology. However, it is important to highlight that the prediction power of the machine learning regression method directly depends on the dataset size, and it requires a large dataset to be employed for modelling. Therefore, the modelling work of this study can only be used for the prediction of Pseudomonas spp. in specific food products (beef, pork, and poultry) and culture medium with certain conditions where a large dataset is available.
Predictive modelling for the growth kinetics of Pseudomonas spp. on button mushroom (Agaricus bisporus) under isothermal and non-isothermal conditions
(Elsevier, 2020) Tarlak, Fatih; Ozdemir, Murat; Melikoglu, Mehmet
Baranyi model was fitted to experimental growth data of Pseudomonas spp. on the button mushrooms (Agaricus bisporus) stored at different isothermal conditions (4, 12, 20 and 28 degrees C), and the kinetic growth parameters of Pseudomonas spp. on the button mushrooms were obtained. The goodness of fit of the Baranyi model was evaluated by considering the root mean squared error (RMSE) and the adjusted coefficient of determination (adjusted-R-2). The Baranyi model gave RMSE values lower than 0.193 and adjusted-R-2 values higher than 0.975 for all isothermal storage temperatures. The maximum specific growth rate (mu(max)) was described as a function of temperature using secondary models namely, Ratkowsky and Arrhenius models. The Ratkowsky model described the temperature dependence of mu(max) better than the Arrhenius model. Therefore, the differential form of the Baranyi model was merged with the Ratkowsky model, and solved numerically using the fourth-order Runge-Kutta method to predict the concentration of Pseudomonas spp. populations on button mushrooms under non-isothermal conditions in which they are frequently subjected to during storage, delivery and retail marketing. The validation performance of the dynamic model used was assessed by considering bias (B-f) and accuracy (A(f)) factors which were found to be 0.998 and 1.016, respectively. The dynamic model developed also exhibited quite small mean deviation (MD) and mean absolute deviation (MAD) values being - 0.013 and 0.126 log CFU/g, respectively. The modelling approach used in this work could be an alternative to traditional enumeration techniques to determine the number of Pseudomonas spp. on mushrooms as a function of temperature and time.

Yazar "Tarlak, Fatih" seçeneğine göre listele

Sayfa Başına Sonuç

Sıralama seçenekleri