Search

Copyright of Turkish Journal of Remote Sensing & GIS is the property of Halil Akinci and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Duzce University Journal of Science & Technology is the property of Duzce University Journal of Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Machine Learning (ML) has become widespread in the food industry and can be seen as a great opportunity to deal with the various challenges of the field both in the present and near future. In this paper, we analyzed 91 research studies that used at least two ML algorithms and compared them in terms of various performance metrics. China and USA are the leading countries with the most published studies. We discovered that Support Vector Machine (SVM) and Random Forest outperformed other ML algorithms, and accuracy is the most used performance metric.

The global impact of the epidemic COVID-19 has done great damage to air transport. Demand for airline transportation has declined for reasons such as quarantine practices by countries, curfews, the economic recession, and the transfer of meetings to digital platforms. This situation has also led to a change in individuals' preferences for air transport. The most striking change in air transport is the tendency of individuals to private air transport privately to minimize the health risks that may arise from personal contacts. Individuals who avoid commercial air transport where public transportation is has transitioned private air transport. For these reasons, an forecasting study was conducted in this study so that a private airline company can provide accurate flight plans in the future. For the forecast study, the number of aircraft types for 2022 was determined by obtaining data on the number of aircraft by passenger capacity, the number of flights, and the number of passengers for 2019-2021 from the airline company. In the forecasting study, the models with the highest accuracy value were selected from the machine learning models. The results provided important information about the company's future fleet planning.

Tek eksenli sıkışma dayanımı (UCS) mühendislik projelerinde en önemli tasarım parametrelerinden biri olup; bir çok projede ve sınıflama sistemlerinde doğrudan kullanılan bir parametredir. UCS’nin elde edilmesindeki güçlükler göz önüne alındığında; makine öğrenimi temelli yaklaşımlar ile tahmin edilmesi dikkat çekmektedir. Çalışma kapsamında bazalt bloklarından alınan 137 adet karot örneği üzerinde gerçekleştirilen laboratuvar deney sonuçları kullanılarak iki ayrı model elde edilmiştir. Bu modellerde görünür gözeneklilik (n), p dalga hızı (Vp) ve birim hacim ağırlık (n) değerleri girdi parametreleri olup; makine öğrenimi yöntemleri ile UCS tahmin edilmeye çalışılmıştır. Bu amaçla; Gauss Süreç Regresyonu (GSR), Destek Vektör Makineleri (DVM) ve Ağaç Toplulukları Yöntemleri (AT) olmak üzere üç farklı makine öğrenimi yöntemi kullanılmıştır. İki ayrı modele ait beş farklı veri seti için uygulanan üç ayrı makine öğrenimi yönteminin performanslarının değerlendirmesinde R2 (determinasyon katsayısı), RMSE (kök ortalama kare hata), MSE (ortalama kare hata) ve MAE (ortalama mutlak hata) performans indisleri kullanılmıştır. Buna göre; genel olarak üç ayrı makine öğrenimi yönteminin de UCS’ nin tahmininde başarılı olduğu değerlendirilmiş olmakla birlikte AT yönteminin genel olarak daha yüksek tahmin performansı verdiği belirlenmiştir.

Anxiety affects productivity and quality of life as well as human abilities and behaviors. It can be considered the main cause of depression and suicide. Clinicians today use specific criteria to diagnose anxiety disorders. There is a need for reliable, non-invasive techniques that fulfil the complex task of detecting anxiety. This study aimed to classify binary and quadruple classes with fewer EEG channels and features by analyzing electroencephalography (EEG) signals. A DASPS database containing 14-channel EEG signals from 23 individuals was used. Using EEGLAB, 4 channels were selected from 14 channels. The wavelet transform is used for feature extraction. The MATLAB Classification learner toolbox contained eight methods for classification. The highest accuracy performances were obtained with the Decision trees method with an accuracy of 67.1% in binary classification, and with a support vector machine with an accuracy of 58.5% in quadruple classification. [ABSTRACT FROM AUTHOR]

The allocation of limited resources and the fair treatment to patients are among the important bioethical issues addressed in the treatment of coronavirus. This study analyzes the relationships between COVID- 19 recovery rate and ventilator allocation time and demographic variables such as gender, race, age, and insurance status. In this study, the data on coronavirus patients of a hospital in the USA were used. Machine learning methods including Random Forest, Decision Tree, Support Vector Machine and Logistic Regression algorithms were utilized to model the data. As a result, that patients covered by Medicaid insurance, which benefits the poor people in the US, are less likely to survive as a result of coronavirus. This result is important in terms of discussing the relationship between health insurance and bioethics.

Copyright of COMU Journal of Agriculture Faculty / ÇOMÜ Ziraat Fakültesi Dergisi is the property of Canakkale Onsekiz Mart University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Weirs are the oldest and most practical structures used to control, regulate, and measure flow in rivers or open channels. The ratio of the actual discharge, smaller than the theoretical discharge due to the separation zone and boundary layer development, to the theoretically discharge is defined as the discharge coefficient (Cd). Cd values are depended on the hydraulic properties of the open channel flow and the type and geometric properties of the weir. In this study, a total of 88 weir head (H0) of the trapezoidal broad crested weir with different upstream and downstream slopes are experimentally measured and Cd values are calculated according to the weir characteristics. Calculated Cd values are estimated by using eight different input combinations with the dimensionless parameters. Three different kernel functions (Linear, Polynomial and Gaussian) of Support Vector Machines method are used. To determine the performance of the models, Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and the coefficient of determination (R2) values are employed. As a result of the study, the Gaussian kernel function is the most successful model and the input combination is H0/L (L as crest height), ɛ (H0/(H0+L)), upstream slope (α) and downstream slope (ꞵ) parameters are found to be most successful model to estimate Cd values.

As stated by the World Health Organization, the occurrence of skin cancer has been increasing in recent years. Between 2 and 3 million non-melanoma skin cancers and at least 132.000 malignant skin cancers occur worldwide each year. Appropriate automatic diagnosis of skin lesions and melanoma recognition can greatly improve the early detection of melanomas. Early diagnosis in skin cancer ensures that patients have the correct diagnosis and treatment. In this study, deep features were extracted from skin lesion images to diagnose whether skin cancer is malignant or not, using cubic-type Support Vector Machine (SVM) classifier and pre-trained Convolutional Neural Network (CNN) based AlexNet and ResNet50 deep architectures, and then combined. Then, effective and distinctive features were selected from these deep features with the ReliefF algorithm. Different classifier algorithms were applied to the combined deep features. Cubic type SVM is used as it gives the best results. In the proposed method, the classification accuracy is 92.41% for the Kaggle dataset and 85.17% for the HAM10000 dataset. In experimental studies, it has been observed that the accuracy score of the proposed model is more successful than other studies. [ABSTRACT FROM AUTHOR]

Recently, there has been an increasing interest in the use of artificial intelligence techniques in different fields. In this work is aimed to use different machine learning algorithms (MLA) to predict the elongation at break from the mechanical properties of cable sheath materials in halogen-free flame retardant (HFFR) cables. In order to be used in the developed prediction models, tensile test was applied to the samples and the percent elongation values were determined. Obtained experimental results were used in different artificial intelligence prediction models. Absolute percentage errors of support vector machine (SVM) and artificial neural network (ANN) methods were obtained at a quite acceptable level with a limited number of data obtained from HFFR cable samples. The estimations obtained by these methods were compared with the data of the estimations obtained by performing regression analysis with the MS Excel program. According to the statistical results, with the use of SVM and ANN in this area, the successful prediction rate was 87.5%, and the average success rate for the predictions made was 92%. The use of MLA in this area will largely end the uncertainty in the trial and error production and reduce the rate of unsuccessful production. [ABSTRACT FROM AUTHOR]

This study aimed to predict the 1 to 2 year future time of the financial failure of 86 manufacturing companies that are operating in Borsa İstanbul. The data comprised of 2010-2012 period, and it depends on 8 quantitative financial variables. Beside 6 variables come from non financial statements. In the study, Artificial Neural Network (NN), Classification and Regression Trees (CART), Support Vector Machine (SVM) and k-Nearest Neighbors (KNN) were used to compare classification performances of related methods. ROC Curve was used to compare the classification performance of the methods. As a result of the analyseis, the overall classification accuracy from the highest to the lowest was SVM (92,31%), CART (88,46%), ANN (84,62%) and KNN (80,77%) 2 years before the financial failure. The overall classification accuracy from the highest to the lowest was CART (96,15%), ANN (92,31%), SVM (80,77%) and KNN (84,62%) 1 year before the financial failure. Return on Equity (ROE) and Return on Assets Ratio (ROA) were found as important variables in the creation of the CART decision tree. The fact that the four models obtained in thise study predicted financial success/failure at a higher rate, and it shows that the models obtained in this study can be included in the models used by relevant people.

Copyright of Ankara University SBF Journal / Ankara Üniversitesi SBF Dergisi is the property of Ankara University SBF Journal and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Objective: Breast cancer; it occurs in the tissue cells of the milk ducts in the breast tissue. The uncontrolled increase in the cells forming the milk ducts is called ductal hyperplasia. It is known that the risk of developing invasive (with a tendency to spread) breast cancer in a woman during her lifetime is 13.3%, and the risk of developing breast cancer increases with age. The Gail model is a well accepted cancer risk assessment model which evaluates the main factors in breast cancer. The aim of this study is to compare machine learning methods in breast cancer risk assessment based on the Gail model. Material and Methods: Firstly, the risk factor was determined by the application of the Gail model into the data set, discrete training test data sets were presented which is 80% train and 20% test. Afterwards, k-nearest neighbor, artificial neural network (ANN), support vector machine (SVM) and naive Bayes (NB) algorithms applied on each set and risk estimation results were compared. Results: Classification performance from the lowest to the highest for 80% training and 20% test data set according to the comparison results is as follows; SVM [area under the curve (AUC)=0.911], NB (AUC=0.939) and ANN (AUC=0.949). Conclusion: Early diagnosis of breast cancer increases the number of possible treatments, the success rate of the treatments and the chance of survival. It has been seen that machine learning algorithms effective in breast cancer risk calculation. [ABSTRACT FROM AUTHOR]

Çevresel etmenler canlı yaşamına doğrudan etki eden birçok doğal afeti tetiklemektedir. Bu afetlerin en önemlilerinden biri de kuraklıktır. Kuraklığın su kaynakları üzerindeki etkisi birçok şekilde canlıları etkilemektedir. Özellikle kuraklığın sebep olduğu, içme suyu ve tarımsal sulama amaçlı kullanılan su kaynaklarında görülen azalmalar, insan yaşamını önemli ölçüde tehdit edebilmektedir. Kuraklık diğer afetler gibi aniden ortaya çıkmadığı için, kuraklık oluşmadan önce tahmin edilip gerekli önlemlerin alınabilmesi imkânı bulunmaktadır. Kuraklık olayının belirlenebilmesi için çeşitli kuraklık indeksleri kullanılmakta ve bu da kuraklığı tahmin edebilme imkânı sunmaktadır. Zaman içerisinde büyük değişiklikler gösteren kuraklık indekslerinin tahmini için birçok araştırma yapılmıştır. Bu çalışmada Kayseri iline ait 116 yıllık Palmer Kuraklık Şiddet İndeksi (PDSI - Palmer Drought Severity Index) değerleri, makine öğrenmesi yöntemleri kullanılarak modellenmiş olup, bir, üç ve altı ay sonraki kuraklık değerleri tahmin edilmiştir. Destek vektör makineleri (SVM) ve K-en yakın komşuluk (KNN) algoritmaları kullanılarak oluşturulan modeller ile yapılan tahminlerin başarı oranı istatistiksel olarak değerlendirilmiştir. Yapılan bu çalışma göstermiştir ki, makine öğrenmesi yöntemleri kuraklık problemlerin çözümüne önemli ölçüde katkı sağlamaktadır.

Çevresel etmenler canlı yaşamına doğrudan etki eden birçok doğal afeti tetiklemektedir. Bu afetlerin en önemlilerinden biri de kuraklıktır. Kuraklığın su kaynakları üzerindeki etkisi birçok şekilde canlıları etkilemektedir. Özellikle kuraklığın sebep olduğu, içme suyu ve tarımsal sulama amaçlı kullanılan su kaynaklarında görülen azalmalar, insan yaşamını önemli ölçüde tehdit edebilmektedir. Kuraklık diğer afetler gibi aniden ortaya çıkmadığı için, kuraklık oluşmadan önce tahmin edilip gerekli önlemlerin alınabilmesi imkânı bulunmaktadır. Kuraklık olayının belirlenebilmesi için çeşitli kuraklık indeksleri kullanılmakta ve bu da kuraklığı tahmin edebilme imkânı sunmaktadır. Zaman içerisinde büyük değişiklikler gösteren kuraklık indekslerinin tahmini için birçok araştırma yapılmıştır. Bu çalışmada Kayseri iline ait 116 yıllık Palmer Kuraklık Şiddet İndeksi (PDSI - Palmer Drought Severity Index) değerleri, makine öğrenmesi yöntemleri kullanılarak modellenmiş olup, bir, üç ve altı ay sonraki kuraklık değerleri tahmin edilmiştir. Destek vektör makineleri (SVM) ve K-en yakın komşuluk (KNN) algoritmaları kullanılarak oluşturulan modeller ile yapılan tahminlerin başarı oranı istatistiksel olarak değerlendirilmiştir. Yapılan bu çalışma göstermiştir ki, makine öğrenmesi yöntemleri kuraklık problemlerin çözümüne önemli ölçüde katkı sağlamaktadır.

To find out which period the historical coins belong to requires a number of scientific procedures that archaeologists or experts can do. These operations can often be time-consuming and demanding operations. From this point on, in this study, the automatically classification of historical coins by using machine learning methods is discussed. Being able to use machine learning methods to classify historical coins can help experts and can become an analysis tool without the need for scientific tests for non-experts. For this purpose, some physical properties of different coins used in Anatolian geography were collected and classified by various machine learning methods named SVM, Random Forest, Bagging, and Decision Trees. Also, two different missing values strategies are deployed in conjunction with each chosen method. Based on our findings, random forest method together with imputing missing values with mean gives an acceptable results with the accuracy rate of %71, although there are some limitations such as high rate of missing values and working with a small dataset.

Differential diagnosis of erythematous squamous diseases is one of the important problems in dermatology. They all share the clinical picture of erythema and scaling among each other, with little difference. The diseases included in this group can be classified as psoriasis, seborrheic dermatitis, lichen planus, pityriasis rosea, chronic dermatitis and pityriasis rubra pilaris. Biopsy for diagnosis, but unfortunately these diseases also share many histopathological features. Technologies related to some other technologies have found wide application in the biomedical field. With the use of computer technologies in medical devices, more sensitive, faster, faster, cutting devices are developed. Therefore, it has been investigated how effective machine learning algorithms are in classifying and predicting skin diseases. In this study, skin tissue samples consisting of 33 attributes belonging to 366 patients, Support Vector Machines (SVM), Ensemble Learning Algorithms (ELA), Decision Trees (DT), k-Nearest Neighborhood (k-NN) were classified with algorithms and the highest knowledge information was recorded. Accordingly, the effects related to the separation and classification of skin diseases have been investigated. SVM has achieved an accuracy of 99.73% which is higher than all the previous studies on this dataset. [ABSTRACT FROM AUTHOR]

Using of solar energy technologies have been developing and increasing in many countries in recent years to meet the energy demand and to provide sustainable energy source. In order to use the developing technologies efficiently, it is necessary to obtain energy source data, and this will ensure that the investments made in this way are more effective. Due to the insufficiency of solar energy measurement devices and the replacement or repair costs of equipment, it is necessary to use the data obtained from meteorological stations as input parameters in the literature and to calculate the solar adiation data with artificial intelligence methods. In this study, the monthly mean solar radiations of Adıyaman and Gaziantep stations are predicted using different input parameters, such as temperature, humidity, mean pressure, wind, number of clear days in a month, and month number. Three different kernel functions (Gaussian, Linear and Polynomial) of support vector machine are used to estimate the average monthly solar radiation. Coefficient of Determination (R²), Root Mean Square Error (RMSE) Mean Absolute Percentage Error (MAPE) Nash-Sutcliffe model efficiency coefficient (NSE) and Percent Bias (PBIAS) parameters are used to determine the performance of selected kernel functions. As a result of the study, it is seen that the Gaussian and Polynomial kernel functions of support vector machines generally show successful output to determine the solar radiation. In addition, it is determined that the models that used mean temperature and mean pressure as input parameters improve the estimation performance. [ABSTRACT FROM AUTHOR]

Bu çalışma BIST 100 borsa endeksinin negatif ve pozitif yönlü hareketlerinin tahmin edilmesini konu edinmektedir. Yapay sinir ağı, destek vektör makinesi ve naive Bayes algoritmasının tahmin performansları karşılaştırılmaktadır. Analizler iki aşamalı olarak yapılmaktadır. Birinci aşamada tahmin modellerinde girdi olarak kullanılacak dokuz adet teknik gösterge, borsa endeksi açılış, kapanış, en yüksek ve en düşük fiyatlar, kullanılarak hesaplanmakta ve sürekli olan bu teknik göstergeler barındırdıkları trende göre kategorize edilerek yeni bir veri seti oluşturulmaktadır. İkinci aşamada ise, trend belirleyici veri seti girdi olarak kullanılmakta ve seçilen üç makine öğrenme algoritması kullanılarak tahminler yapılmaktadır. BIST 100 veri seti 2009-2018 Aralığını kapsayan günlük kapanış fiyatlarını içermektedir. Analizlerle, destek vektör makineleri algoritmasının en iyi sınıflandırıcı olduğu sonucuna ulaşılmıştır. Ayrıca, daha önceki benzer çalışmalarla karşılaştırmalar yapılarak gerek kullanılan veri seti gerekse tahmin modellerinin etkileri tartışılmaktadır.

Automatic recognition of the age and gender groups of the speakers is an important research topic and is used for different purposes in many fields, especially in call centers. In this study, the effects of logarithmic energy and delta parameters added to Mel Frequency Cepstral Coefficients (MFCC) on automatic age and gender recognition were investigated. After transforming the MFCC features extracted from speech signals into Gaussian Mixture Model (GMM) supervectors, they were applied to the Support Vector Machine (DVM) and the age and gender groups of the speakers were decided at the end of the optimization process. In the study, besides the parameters added to MFCC, the effects of MFCC number and GMM component number on success were also investigated. MFCC number was changed between 8 and 20 and GMM component number was changed between 32 and 256 and tests were performed on the system. In tests performed with 1388 speeches of 299 speakers in the development section of aGender database, the highest classification rate was calculated as 60.23% by adding logarithmic energy, delta and delta-delta parameters to 12 cepstral coefficients. In the study, the optimum GMM component number was determined as 128, while the effects of logarithmic energy, delta and delta-delta parameters on success were 1.17%, 3.24% and 4.61%, respectively. [ABSTRACT FROM AUTHOR]

This study aimed to predict the 1 to 2 year future time of the financial failure of 86 manufacturing companies that are operating in Borsa İstanbul. The data comprised of 2010-2012 period, and it depends on 8 quantitative financial variables. Beside 6 variables come from non financial statements. In the study, Artificial Neural Network (NN), Classification and Regression Trees (CART), Support Vector Machine (SVM) and k-Nearest Neighbors (KNN) were used to compare classification performances of related methods. ROC Curve was used to compare the classification performance of the methods. As a result of the analyseis, the overall classification accuracy from the highest to the lowest was SVM (92,31%), CART (88,46%), ANN (84,62%) and KNN (80,77%) 2 years before the financial failure. The overall classification accuracy from the highest to the lowest was CART (96,15%), ANN (92,31%), SVM (80,77%) and KNN (84,62%) 1 year before the financial failure. Return on Equity (ROE) and Return on Assets Ratio (ROA) were found as important variables in the creation of the CART decision tree. The fact that the four models obtained in thise study predicted financial success/failure at a higher rate, and it shows that the models obtained in this study can be included in the models used by relevant people. [ABSTRACT FROM AUTHOR]

Copyright of COMU Journal of Agriculture Faculty / ÇOMÜ Ziraat Fakültesi Dergisi is the property of Canakkale Onsekiz Mart University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Beyin bilgisayar arayüzleri (BBA) sadece beyinde üretilen işaretleri kullanarak çeşitli elektronik cihazları kullanmayı olanaklı hale getiren sistemlerdir. Bu sistemlerin yüksek başarımlı olabilmesi için bu işaretlerden çıkarılan öznitelik yöntemleri ve bu işaretlere uygulanan sınıflandırıcı yöntemleri önemlidir. Bu çalışma ile motor hayaline dair kaydedilen EEG tabanlı BBA işaretlerinden yüksek sınıflandırma doğruluğu elde edebilmek için işaretlerin etkin zaman dilimlerinden çıkarılmış özniteliklerle sınıflandırma doğruluğunun artırılmasına yönelik bir yöntem önerilmiştir. Öznitelikler, etkin zaman dilimleri belirlenen EEG işaretlerine Hilbert Dönüşümü’nün uygulanması ve işaretin türevlerinin ortalamasının alınmasıyla elde edilmiştir. BCI Competition 2003 yarışmasında kullanıma sunulmuş 2-sınıflı motor hareketi hayaline dayalı Data Set Ia isimli veri kümesinden çıkarılan öznitelikler destek vektör makineleri, k-en yakın komşuluk ve doğrusal ayrım analizi ile test edilerek performans karşılaştırması yapılmıştır. Destek vektör makineleri ile test veri kümesi üzerinde %91.46 oranında yüksek bir sınıflandırma doğruluğu elde edilmiştir. Bu sınıflandırma doğruluğu EEG işaretinin bir denemesine ait tüm örneklemelerin kullanılması durumunda elde edilen sınıflandırma doğruluğundan %17.40 daha yüksektir. Elde edilen sonuçlar, önerilen yöntemin belirlenen öznitelik çıkarma yöntemi ve destek vektör makinaları sınıflandırıcısıyla birlikte EEG işaretlerinden elde edilen sınıflandırma doğruluğunu dikkat çekici miktarda arttırdığını ve hesaplama karmaşıklığını ise azalttığını göstermiştir.

Beyin bilgisayar arayüzleri (BBA) sadece beyinde üretilen işaretleri kullanarak çeşitli elektronik cihazları kullanmayı olanaklı hale getiren sistemlerdir. Bu sistemlerin yüksek başarımlı olabilmesi için bu işaretlerden çıkarılan öznitelik yöntemleri ve bu işaretlere uygulanan sınıflandırıcı yöntemleri önemlidir. Bu çalışma ile motor hayaline dair kaydedilen EEG tabanlı BBA işaretlerinden yüksek sınıflandırma doğruluğu elde edebilmek için işaretlerin etkin zaman dilimlerinden çıkarılmış özniteliklerle sınıflandırma doğruluğunun artırılmasına yönelik bir yöntem önerilmiştir. Öznitelikler, etkin zaman dilimleri belirlenen EEG işaretlerine Hilbert Dönüşümü’nün uygulanması ve işaretin türevlerinin ortalamasının alınmasıyla elde edilmiştir. BCI Competition 2003 yarışmasında kullanıma sunulmuş 2-sınıflı motor hareketi hayaline dayalı Data Set Ia isimli veri kümesinden çıkarılan öznitelikler destek vektör makineleri, k-en yakın komşuluk ve doğrusal ayrım analizi ile test edilerek performans karşılaştırması yapılmıştır. Destek vektör makineleri ile test veri kümesi üzerinde %91.46 oranında yüksek bir sınıflandırma doğruluğu elde edilmiştir. Bu sınıflandırma doğruluğu EEG işaretinin bir denemesine ait tüm örneklemelerin kullanılması durumunda elde edilen sınıflandırma doğruluğundan %17.40 daha yüksektir. Elde edilen sonuçlar, önerilen yöntemin belirlenen öznitelik çıkarma yöntemi ve destek vektör makinaları sınıflandırıcısıyla birlikte EEG işaretlerinden elde edilen sınıflandırma doğruluğunu dikkat çekici miktarda arttırdığını ve hesaplama karmaşıklığını ise azalttığını göstermiştir.

Copyright of International Journal of InformaticsTechnologies is the property of Institute of Informatics, Gazi University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The main problem in the classification problems encountered with gene samples is that the dimension of the data is high although the sample size is small. In such problems, the classifier to be used must be a classifier that allows the processing of high dimensional data and extracts maximum information from a small number of samples at hand. In this context, a classification methodology has been developed, which first transforms the problem of binary or multiple classification into separate pair-wise classification problems. To this end, an online classifier has been adapted to solve pair-wise binary classification problems. The resulting classifier performed better on most of the real problems compared to other popular classifiers.

Environmental factors, which directly affect the living beings cause the formation of many natural disasters. One of the most important of these disasters is drought. The effect of the drought on the water resources also affects many things in the way of living life. From the point of human life, diminution in water resources, may pose a significant threat. Drought does not appear suddenly, hence it is possible to predict and take necessary measures before it exists. In order to predict the drought, various drought indices are used to determine the drought phenomenon. A great deal of research has been made to estimate the drought values that have changed dramatically so far. In this study, the 116 - year Palmer Drought Severity Index (PDSI) values of Kayseri province were modeled using machine learning methods in order to predict future PDSI values. In this context, one, three and six months period of drought values were predicted. The success rate of the predictions constructed using support vector machines (SVM) and K-nearest neighbors (KNN) algorithms was evaluated statistically. This study indicates that machine learning methods provide a significant contribution to the solution of hydrological problems. [ABSTRACT FROM AUTHOR]

In this study, a classification structure consisting of five processing stages was organized for the diagnosis of ALS and Myopathic diseases, the most common types of neuromuscular diseases. EMG (Electromyogram) signals have been passed through pre-processing, division, clustering, and feature extraction steps before being classified. Hybrid clustering method is used in clustering phase. Afterwards, feature vectors intime and frequency domains and their different combinations of multiple feature vectors (a total of 25 feature vectors) are used. In the next step, data are classified by Support Vector Machine (DVM), K-Nearest Neighbor (K-EYK) algorithm and Discriminant Analysis (DA) algorithms. Verification is used as a measure of cross-validation method. Cross-validation of the k-value of 10 is selected. Experimental results show that the proposed feature vectors are more successful than the single feature vectors of multiple feature vectors. When usedin multiple feature vectors; SVM classifier, has classified the EMG signals withhigher accuracy in according to the K-NN and DA classifiers. Total accuracy is97.39% for ALS and 86.74% for the myogenic. It is understood with this study; the classification performance depends on a high degree of feature vectors of interclass separability. [ABSTRACT FROM AUTHOR]

In this study, wind energy potential of Sinop and Adıyaman provinces in different regions of Turkey were analyzed statistically, based on the hourly measured data by Directorate of State Meteorological Station in 2008-2017 years. During the statistical analysis, average wind speed, standard deviation of wind speed, maximum wind speed and wind power density were determined. The Weibull distribution function was used to the distribution of wind speed and determination of wind power intensity. For the power density values obtained as a result of the study, a predictive model was established with the support vector machine (SVM) regression. Polynomial kernel, normalized polynomial kernel, radial basis function (RBF) kernel and Pearson universal kernel VII (PUK) models were used in SVM regression. Mean absolute error (MAE), root mean square error (RMSE), relative absolute error (RAE), and root relative square error (RRSE) error analyzes were performed for SVM regression estimates. The best estimation of wind power density predictive models generated by 4 different kernel functions using SVM regression was shown to belong to the polynomial kernel. [ABSTRACT FROM AUTHOR]

Copyright of SDU Journal of Engineering Sciences & Design / Mühendislik Bilimleri ve Tasarım Dergisi is the property of Journal of Engineering Sciences & Design and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In addition to the uncertainty and chaotic movements of the financial time series, the nonlinear dynamic structure makes the forecasts very difficult. The fact that the stock market index are affected by the political changes, the general outlook of the economy, the investors' expectations and investment preferences, and the movements of other indexes, make the index estimates quite difficult but attractive. It is known that the machine learning algorithms are successful in estimating stock index movements and their future values. In this study, the problem of forecasting the direction of BIST 100 index movements is discussed. Three different machine learning algorithms, artificial neural networks, support vector machines and naïve Bayes classifier were used and their performances were compared. Ten technical indicators were used as inputs for the models. The data set consists of ten-year daily closing price values covering the 2009-2018 period. Analysis results show that the models can be used to capture stock market index movements, whereas artificial neural network algorithm is a better classifier. [ABSTRACT FROM AUTHOR]

Brain computer interfaces (BCI) are systems which make it possible to use various electronic devices using only the signals produced in the brain. In order to ensure high performance of these systems, feature methods extracted from these signals and classifier methods applied to these signals are important. With this study, we proposed a method to obtain high classification accuracy from EEG based BBA signals recorded on the motor imaginary with the extracted features in the active time segments. Features were obtained by applying the Hilbert Transform to the active time segments selected EEG signs and calculating the average of the derivatives of the signs. Features extracted from two-class motor imaginary Data Set Ia (Presented at the BCI Competition 2003 competition) were analyzed by support vector machines, k-nearest neighborhood and linear discriminant analysis. Then the performance of the classifiers was compared. A high classification accuracy of 91.12% is calculated on the test dataset with support vector machines. This classification accuracy is 17.06% higher than the classification accuracy obtained in the case of using all samples of a trial of the EEG signal. As a result, the proposed method increased the accuracy of classification in a remarkable amount and reduced computational complexity with the feature extraction methods and support vector machine classifier. [ABSTRACT FROM AUTHOR]

Objective: The primary aim of this study is to examine and compare the classification performance of support vector machine mo dels generated by various core functions used to classify diabetes mellitus in acute coronary syndrome patients. The secondary aim is to optimize the parameters of the various kernel functions which are used for constructing the support vector machine mo del and to achieve the best classificat ion performance. Material and Method: The data examined in this study were selected retrospectively from the database developed for Inonu University Turgut Ozal Medical Center Cardiology Department. The study included type 2 diabetes mellitus and various demographic and clinical variables in acute coronary syndrome patients. The Support Vector Machine model was used to classify type 2 diabetes mellitus in acute coronary syndrome pat ients. The related models are constructed by ANOVA radial basis funct ion, bessel, linear, Gaussian radial basis funct ion, laplace, polynomial and sigmoid kernel funct ions. Results: The best classification performance was obtained by Support Vector Machine model const ructed by laplace kernel funct ion based on the results of performance metrics. The accuracy, area under ROC curve, sensit ivity and specificity met rics with 95% CI were calc ulated as; 0.9804 (0.9716 - 0.987), 0.9332 (0.9096 - 0.9567), 0.9999 (0.9791 - 1.000) and 0.9776 (0.9675 - 0.9852), respect ively. Conclusion: When the performance metrics were taken into account, the best classification per formance was achieved from the Laplace Support Vector Machine model. In subsequent studies, the use of Support Vector Machine models with different kernel functions and other machine learning or data mining algorithms in different clinical trials may improve the classification success of the diseases. [ABSTRACT FROM AUTHOR]

Matematik mühendislik eğitimindeki önemli konulardan biridir. Mühendislik fakültelerine temel matematik bilgi ve becerisi yetersiz olarak gelen öğrenciler, matematik dersinde riskli öğrenciler olarak sınıflandırılmaktadır. Bu çalışmanın amacı, Destek Vektör Makineleri yöntemini kullanarak riskli öğrencileri tahmin etmektir. 2007 yılında Pamukkale Üniversitesi mühendislik fakültesine başlayan 434 öğrencinin üniversiteye giriş sınavı sonuçlarına ait verilerin kullanıldığı araştırma sonuçlarına göre; öğrencilerin matematik, fen bilimleri, Türkçe testlerinin sonuçları ile lise mezuniyet başarı puanlarının, matematik I dersindeki başarılarını tahminde önemli rol oynadığı bulunmuştur. 289 öğrencinin verisi ile makine eğitilmiş, 145 öğrencinin verisi ise test için kullanılmış ve matematik I dersinden geçen öğrencilerin % 86'sı doğru olarak tahmin edilmiştir.

Matematik mühendislik eğitimindeki önemli konulardan biridir. Mühendislik fakültelerine temel matematik bilgi ve becerisi yetersiz olarak gelen öğrenciler, matematik dersinde riskli öğrenciler olarak sınıflandırılmaktadır. Bu çalışmanın amacı, Destek Vektör Makineleri yöntemini kullanarak riskli öğrencileri tahmin etmektir. 2007 yılında Pamukkale Üniversitesi mühendislik fakültesine başlayan 434 öğrencinin üniversiteye giriş sınavı sonuçlarına ait verilerin kullanıldığı araştırma sonuçlarına göre; öğrencilerin matematik, fen bilimleri, Türkçe testlerinin sonuçları ile lise mezuniyet başarı puanlarının, matematik I dersindeki başarılarını tahminde önemli rol oynadığı bulunmuştur. 289 öğrencinin verisi ile makine eğitilmiş, 145 öğrencinin verisi ise test için kullanılmış ve matematik I dersinden geçen öğrencilerin % 86'sı doğru olarak tahmin edilmiştir.

In this study, a system classifying speakers according to their age and/or genders is proposed. In this system phone conversations including mobile calls that took place indoor or outdoor are used as inputs. It is aimed to classify the speakers according to their genders into three classes as male, female and child, according to their ages into four classes as child, youth, adult and senior, and finally according to both gender and age into seven classes. For this aim, GMM models that are created with MFCC coefficients obtained by the voiced parts of the conversations are transformed into supervectors. These supervectors are applied to SVM classifier. Signal energy is used for determining the voiced parts of conversations. For the training of GMM models, the adaptation approach of UBM is preferred. Also, by testing GMM models that are created with different number of components and different length conversations, the impact of GMM components number and speech duration on the age and gender identification is investigated. At the end of these tests, the highest classification success rates are obtained by modeling 16-second speeches with 64-component GMMs. The rates obtained from these tests are measured as 92.42% for gender category, 60.10% for age category and 60.02% for age&gender category. [ABSTRACT FROM AUTHOR]