In this study, the freshness levels of boiled chicken eggs were determined using an electronic nose and machine learning techniques. Eggs were boiled and stored under refrigerator conditions (3±1ºC) from day 0 to day 6. Each storage day, eggs were divided into two groups based on cooling methods: quick-cooled and fast-cooled. Sensor readings were taken using an electronic nose, and image changes from 110 daily image files were processed with a machine learning program. With 85% of the image data used for training and 15% for testing, a classification accuracy of over 98% was achieved. The results showed that egg white solidified in more than 4 minutes and yolk solidified in 11 minutes. Fast-cooled eggs exhibited significantly lower odor levels, indicating superior freshness. This study demonstrates the effectiveness of electronic nose and machine learning systems in accurately determining the freshness of boiled eggs.

In this study, the freshness levels of boiled chicken eggs were determined using an electronic nose and machine learning techniques. Eggs were boiled and stored under refrigerator conditions (3±1ºC) from day 0 to day 6. Each storage day, eggs were divided into two groups based on cooling methods: quick-cooled and fast-cooled. Sensor readings were taken using an electronic nose, and image changes from 110 daily image files were processed with a machine learning program. With 85% of the image data used for training and 15% for testing, a classification accuracy of over 98% was achieved. The results showed that egg white solidified in more than 4 minutes and yolk solidified in 11 minutes. Fast-cooled eggs exhibited significantly lower odor levels, indicating superior freshness. This study demonstrates the effectiveness of electronic nose and machine learning systems in accurately determining the freshness of boiled eggs.

The objective of this study was to compare predictive performances of four machine learning (ML) models: Support Vector Machines with Radial Basis Function Kernel (SVMR), Classification and Regression Trees (CART), Random Forest (RF) and Model Average Neural Networks (MANN) to predict live weight from morphological measurements of Norduz sheep (n=93). Seven morphological measurements; chest girth (CG), chest width (CW), chest depth (CD), height at withers (HW), body length (BL), heigth at rump (HR) and rump width (RW) were used to predict live weigth (LW) of Norduz sheep. All morphological measurements were positively correlated to LW. Live weight had the highest correlation with CG and the lowest correlation with HR. Initially, highly correlated predictors were removed from the data set. The remaining predictors were then subjected to variable selection procedures using the Boruta algorithm. The results of Boruta confirmed the importance of the four predictors HW, BL, CW, and CD. However, HR confirmed to be unimportant was excluded from the dataset. The ML models were trained on selected predictors. The results showed that the prediction performance validated using the test dataset indicated that RF had the lowest values of Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Mean Absolute Percent Error (MAPE). The permutation-based variable importance scores indicate that CW and CD were the most important variables in predicting LW. The actual LW had the highest significant positive correlations with the values predicted by SVMR and RF, and followed by ANN and CART models respectively. There were no differences between the means of actual and predicted LWs by machine learning models. The fact that the models generalized well on the testing data sets indicates that machine learning algorithms have valid predictive patterns and are effective methods in LW weight of Norduz sheep. Considering runtime of the models, although the CART model had the lowest computational cost, it had the worst performance. The MANN algorithm, on the other hand, required a longer runtime to process the same dataset.

Erzincan Tulum Cheese (ETC) holds a significant place among the most popular cheeses in Türkiye. It has been awarded Protected Geographical Indication status, which restricts the allowable milk species, its production area, and specific sheep breed used in its production. Mineral content and antioxidant activity of ETC were aimed to be predicted using conventional FT-NIR and a portable FT-MIR spectrometer combined with partial least square regression (PLSR) and machine learning algorithms based on conditional entropy. Seventy ETC samples were analyzed for their mineral (Al, Ca, Cr, Cu, Fe, K, Mg, Mn, Na, and P) content using ICP-MS. The samples' antioxidant activity was measured using the DPPH•+ scavenging activity method. PLSR combined with FT-NIR spectral data correlated with antioxidant activity (r=0.89) and minerals (as low as r=0.83) except for Cr and Fe. FT-MIR data provided a good correlation for minerals (as low as r=0.82) except for Cr and Mn and a moderate correlation with antioxidant activity (r=0.64). Information theory was applied to select wavenumbers used in machine learning algorithms, and better results were obtained compared to PLSR. Overall, FT-NIR and FT-MIR spectroscopy provided rapid (~ 1 min), non-destructive, sensitive, and reliable output for mineral and antioxidant activity predictions in commercial cheese samples.

This is study is based with the support of RS-GIS technology on the land use of Samsun Center, as well as the coastal districts of Ilkadım,Atakum,Bafra Plain, through the processing and interpretation of satellite images in the summer months of 2000,2010,2020. Spatial and temporal variability properties of LU/LC were determined using MLC algorithm, controlled classification approach. The predictive values of the LU/LC change that will occur in 2030, calculated with the MLP‑ANN model based on Machine Learning algorithms and mapped with the QGIS 3.16 program. To determine the accuracy coefficient of the model, 2020 LU/LC simulation performed using the transition potential matrix of 2000 and 2010 LU/LC data. The results of simulation were compared the data of land use land cover with the 2020 to evaluate the accuracy of the simulation model. The model of MLP‑ANN provided an accuracy of 72% based on the kappa fit index. According to MLP‑ANN model 2030 results were an increase of 73.33 km² in built up areas, an increase of 56.89 km² in bare areas, and a decrease of 129.66 km² in green areas. It provided a reference basis for future Samsun urban to rural coastline LU planning and management and LU structure optimization.

Fish is regarded as an important protein source in human nutrition due to its high concentration of omega-3 fatty acids In traditional global cuisine, fish holds a prominent position, with seafood restaurants, fish markets, and eateries serving as popular venues for fish consumption. However, it is imperative to preserve fish freshness as improper storage can lead to rapid spoilage, posing risks of potential foodborne illnesses. To address this concern, artificial intelligence techniques have been utilized to evaluate fish freshness, introducing a deep learning and machine learning approach. Leveraging a dataset of 4476 fish images, this study conducted feature extraction using three transfer learning models (MobileNetV2, Xception, VGG16) and applied four machine learning algorithms (SVM, LR, ANN, RF) for classification. The synergy of Xception and MobileNetV2 with SVM and LR algorithms achieved a 100% success rate, highlighting the effectiveness of machine learning in preventing foodborne illness and preserving the taste and quality of fish products, especially in mass production facilities.

This is study is based with the support of RS-GIS technology on the land use of Samsun Center, as well as the coastal districts of Ilkadım,Atakum,Bafra Plain, through the processing and interpretation of satellite images in the summer months of 2000,2010,2020. Spatial and temporal variability properties of LU/LC were determined using MLC algorithm, controlled classification approach. The predictive values of the LU/LC change that will occur in 2030, calculated with the MLP‑ANN model based on Machine Learning algorithms and mapped with the QGIS 3.16 program. To determine the accuracy coefficient of the model, 2020 LU/LC simulation performed using the transition potential matrix of 2000 and 2010 LU/LC data. The results of simulation were compared the data of land use land cover with the 2020 to evaluate the accuracy of the simulation model. The model of MLP‑ANN provided an accuracy of 72% based on the kappa fit index. According to MLP‑ANN model 2030 results were an increase of 73.33 km² in built up areas, an increase of 56.89 km² in bare areas, and a decrease of 129.66 km² in green areas. It provided a reference basis for future Samsun urban to rural coastline LU planning and management and LU structure optimization.

The objective of this study was to compare predictive performances of four machine learning (ML) models: Support Vector Machines with Radial Basis Function Kernel (SVMR), Classification and Regression Trees (CART), Random Forest (RF) and Model Average Neural Networks (MANN) to predict live weight from morphological measurements of Norduz sheep (n=93). Seven morphological measurements; chest girth (CG), chest width (CW), chest depth (CD), height at withers (HW), body length (BL), heigth at rump (HR) and rump width (RW) were used to predict live weigth (LW) of Norduz sheep. All morphological measurements were positively correlated to LW. Live weight had the highest correlation with CG and the lowest correlation with HR. Initially, highly correlated predictors were removed from the data set. The remaining predictors were then subjected to variable selection procedures using the Boruta algorithm. The results of Boruta confirmed the importance of the four predictors HW, BL, CW, and CD. However, HR confirmed to be unimportant was excluded from the dataset. The ML models were trained on selected predictors. The results showed that the prediction performance validated using the test dataset indicated that RF had the lowest values of Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Mean Absolute Percent Error (MAPE). The permutation-based variable importance scores indicate that CW and CD were the most important variables in predicting LW. The actual LW had the highest significant positive correlations with the values predicted by SVMR and RF, and followed by ANN and CART models respectively. There were no differences between the means of actual and predicted LWs by machine learning models. The fact that the models generalized well on the testing data sets indicates that machine learning algorithms have valid predictive patterns and are effective methods in LW weight of Norduz sheep. Considering runtime of the models, although the CART model had the lowest computational cost, it had the worst performance. The MANN algorithm, on the other hand, required a longer runtime to process the same dataset.

In this study, an image processing algorithm and classification unit were developed to classify the fruits according to their size and color characteristics. For this purpose, a total of 300 fruits (50 fruit samples from each of the Starkrimson Delicious and Golden Delicious apple varieties, Washington Navel and Valencia Midknight orange varieties, Ekmek and Eşme quince varieties) were used in the experiments. The size and color values measured with a caliper and a spectrophotometer were entered in the developed image processing algorithm to determine the success rates of classifying the fruits. The integration of image processing algorithm with the classification unit classified 88%, 100%, 96%, 82%, 86%, respectively. On the other hand, the size and color values read in fruits with the image processing algorithm were evaluated using predictive techniques used in data mining. For this purpose, K Nearest Neighbor (KNN), Decision Tree (DT), Naive Bayes classification and Multilayer Perceptron Neural Network (MLP) algorithms were used. Algorithms were run with 10-fold cross validation method. In the training of artificial classifiers, the success was 93.6% for KNN, 90.3% for DT, 88.3% for Naive Bayes, 92.6% for MLP and 94.3% for RF.

In this study, an image processing algorithm and classification unit were developed to classify the fruits according to their size and color characteristics. For this purpose, a total of 300 fruits (50 fruit samples from each of the Starkrimson Delicious and Golden Delicious apple varieties, Washington Navel and Valencia Midknight orange varieties, Ekmek and Eşme quince varieties) were used in the experiments. The size and color values measured with a caliper and a spectrophotometer were entered in the developed image processing algorithm to determine the success rates of classifying the fruits. The integration of image processing algorithm with the classification unit classified 88%, 100%, 96%, 82%, 86%, respectively. On the other hand, the size and color values read in fruits with the image processing algorithm were evaluated using predictive techniques used in data mining. For this purpose, K Nearest Neighbor (KNN), Decision Tree (DT), Naive Bayes classification and Multilayer Perceptron Neural Network (MLP) algorithms were used. Algorithms were run with 10-fold cross validation method. In the training of artificial classifiers, the success was 93.6% for KNN, 90.3% for DT, 88.3% for Naive Bayes, 92.6% for MLP and 94.3% for RF.