Search

Turkish manufacturing industry is in the top three in terms of occupational accident frequency among sectors. Therefore, there is a need to determine accident cause-effect relationships in order to improve occupational safety and minimize the risks that cause occupational accidents in the manufacturing industry. An integreated data driven approach is proposed to find patterns among occupational accidents in Turkish manufacturing systems. The proposed approach uses multinomial logit model (MLM) and decision tree algorithms, namely C5.0, Classification and Regression Trees (C&RT), The quaternion estimation (QUEST), Chi-square automatic interaction detector (CHAID) ve Random Trees. In this study, 307,590 occupational accidents in the Turkish manufacturing industry between 2013 and 2019 are used. It is found that there is a statistically significant relationship among division, geographical location of the accident, year, deviation, hour day, gender and age for all accidents with injury, death and loss of limb according to the absence of disability. Additionally, division, geographical location of the accident and year are among the top five predictors based on decision tree algorithms. [ABSTRACT FROM AUTHOR]

The secondhand vehicle market in Türkiye has always been active. In the secondhand vehicle market, information such as brand, model, and fuel type can be analyzed, and this information can be made useful. Prices vary according to the various features of the vehicles. Machine learning techniques can predict prices and help users set prices when selling or buying vehicles. Price prediction falls under regression. Since the number of secondhand vehicles is quite high, big data systems are used. Apache Spark and its machine learning library are quite useful for this. Linear regression, decision tree regression, random forest regression, GBT regression, and isotonic regression algorithms are used for price prediction. The random forest algorithm achieved the highest success for the price prediction with 21435.09 RMSE and 0.887 R2 values. As a result of the statistical tests performed to check the significant difference between the RMSE and R2 values obtained with the random forest algorithm and other algorithms, it is concluded that the results obtained with the random forest algorithm are statistically better than other algorithms. The random forest algorithm gives better results because the algorithm performs training over multiple decision trees, its flexibility, and strong hyperparameters. [ABSTRACT FROM AUTHOR]

The concept of functional data, defined by Ramsey and Dalzell (1991), has led to the emergence of the field of functional data analysis, which helps to interpret the structure and relations in the data much more easily, especially by using visual approaches, by transforming the quantitative data observed at discrete points into functional data. The science mapping technique is exploratory research that reveals the development of a field of interest, such as functional data analysis, in the literature. Thus, focal points such as publication type, publisher, author, origin, citation and reference information and keywords can be displayed together over the years. In this study, the Web of Science (WoS) database example is discussed. In the statistical method studies in WoS, 6494 publications related to functional data analysis were examined with the science mapping technique. In the last 32 years, it has been seen that various analysis methods used in connection with functional data have been developed. It has been determined that the application of machine learning techniques to functional data is being examined in the literature, especially as of 2018. In addition, decision trees, one of the supervised machine learning methods, are associated with functional data. [ABSTRACT FROM AUTHOR]

The aim of this study is to measure the attitude of the students studying in the mathematics department about the derivative subject in the analysis courses taught in the department, with the decision tree optimization algorithm. In the study, a derivative attitude scale with a 5-point Likert scale developed by Kara (2014) and optimized by Atasoy and Kara (2021) was used. This scale was applied to 194 students studying in the mathematics department and it was examined whether there was a difference in the opinions of the students who attended or did not attend the reinforcement courses at the high school level. Considering the number of questions that students answered correctly in the OSYM mathematics test, it was seen that they viewed the derivatives in the university analysis course positively. [ABSTRACT FROM AUTHOR]

Nowadays, with the widespread use of e-commerce, online sales are becoming more and more important for businesses. This increase in the use of e-commerce brings with it the question of what factors may affect the use of e-commerce by consumers. In this study, the features that affect the e-commerce usage frequency of individuals in Turkey are analyzed with decision trees and Support Vector Machine (SVM) algorithms and interpreted. In addition, a classification model has been designed to predict e-commerce usage frequency of individuals by using variables that affect e-commerce usage. According to the results of the study, the most important variable which affects the e-commerce usage frequency in CHAID, C&R Tree and QUEST algorithms has been determined as the number of purchases of goods or services over the internet for private use in the last three months. While variable of age has been the second most important variable in CHAID and QUEST algorithms, the variable of sending/receiving e-mails for special purposes (including mobile applications) in the last three months has been the second most important variable in the C&R Tree. For e-commerce usage frequency estimation, the CHAID algorithm has given the best classification result with an accuracy rate of 86.19%. [ABSTRACT FROM AUTHOR]

With digitalization, the world of crime has also become digital and the number of crimes committed over the internet is increasing day by day. Cybercriminals and attackers use secret networks on the Internet, called the Dark Web, to hide their identities and provide encrypted communication. Darknets have different and special access methods than normal internet infrastructure. All access to these networks is suspect and needs to be investigated. Because the Darknet provides encrypted communication, it is difficult to detect and classify with today's security tools. In this study, only the statistical information of packets was analyzed using machine learning approach without deciphering encrypted network traffic. CICDarknet2020 dataset was used and a detailed experimental study including K Nearest Neighbor, Logistic Regression, Random Forest, SVM, Decision Tree, Gaussian Naive Bayes, Linear Discriminatory Analysis, Gradient Boosting, Extra Tree and XGBoost algorithms was carried out for packet analysis. In experimental studies, it has been observed that the Decision Tree algorithm has the highest classification success with an accuracy rate of 93.32%. [ABSTRACT FROM AUTHOR]

The increasing urbanization and rapid economic development have led to growing housing markets. This has resulted in a rising demand for housing due to migration towards cities and the acceptance of housing as both an investment and a wealth protection tool. On the other hand, housing prices are affected by many factors, thereby studies on housing price forecasts are carried out within the scope of different models and various methods. In line with the significant importance of housing markets as one of Turkey's economic growth tools, this study includes the forecast of the classification of 4-bedroom (3+1) housing prices in the central district of Kütahya province. The dataset used includes various criteria that affect housing prices, and different machine learning classification algorithms that are utilized. As a result, it is seen that the accuracy rates of the k-Nearest Neighbor (k-NN), Decision Trees (DT) and Random Forest (RF) algorithms were above 60%. As the main performance criterion, the AUC scores yielded that the most successful classification forecast is acquired by the RF algorithm, which is followed by the k-NN. In the context of the variables and data set included in this study, it was seen that successful outcomes in the classification forecasts of 4-bedroom apartments’ housing prices in the city center of Kütahya were obtained by these two methods. [ABSTRACT FROM AUTHOR]

In sustainable agriculture, it is of great importance to decide which type of seed should be grown in which soil type. Determining the dried bean type to be planted is a costly process that requires experience, knowledge and experts in the field. Decision support systems developed for the correct classification of dry beans are of great importance at this point. In this study, a hybrid model was proposed for the classification of dry bean seeds using Artificial Neural Network (ANN) and Particle Swarm Optimization (PSO) algorithm. Thanks to the proposed hybrid model, the hyper-parameters in the ANN network structure have been optimized. The ANN network structure established with the hyper parameters obtained as a result of the optimization process was used in the multi-class classification of dry beans. The results obtained with the hybrid model are controversially compared with the classification results using the Decision Tree, K Nearest Neighbor Algorithm. As a result of the experimental studies, successful results were obtained with 90% accuracy in the classification of dry bean seeds with the hybrid model. [ABSTRACT FROM AUTHOR]

In recent years, gender detection has been an important problem in speech analysis, with many different applications. Different features of sound data such as pitch, median, and frequency are used for gender detection. In this study, a feature selection method based on metaheuristic optimization algorithms is proposed for gender detection from voice data. In the proposed method, the feature set representing the voice data in the most appropriate way is selected with optimization algorithms, and gender detection has been implemented with artificial intelligence algorithms using the obtained features. Nature-inspired metaheuristic optimization algorithms, which are capable of solving complex problems, are used to select features from voice data. Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Salp Swarm Algorithm (SSA), and Whale Optimization Algorithm (WOA) have been modeled for the first time for feature selection from voice data. A publicly accessible data set has been used to measure the efficiency of metaheuristic optimization algorithms. The performances of PSO, ACO, SSA, and WOA for feature selection have been compared in terms of three different criteria: fitness function value, accuracy value, and the number of selected features. After feature selection with metaheuristic optimization algorithms, Naive Bayes and Decision Tree algorithms have been applied to the new data sets obtained and the original data set. As a result of the analysis, it was observed that this method which uses metaheuristic optimization algorithms for feature selection increased the success rate in the results obtained with Naive Bayes and Decision Tree algorithms. [ABSTRACT FROM AUTHOR]

In the study, it was aimed to examine the relationship between the psychological resilience of mothers and their children's social skills and the factors affecting social skills. For this purpose, the research was designed as a study in which the predictive correlational research model and the decision tree from the classification models were used. The sample group consisted of 371 children aged 48-72 months and their mothers. "General Information Form", "Adult Resilience Scale" and "Preschool Social Skills Assessment Scale" were used as data collection tools. Pearson Correlation Coefficient, simple linear regression and CHAID algorithm were used in the analysis of the data. As a result of the research, it was seen that there was a moderately positive and significant relationship between the psychological resilience of the mothers and the social skills of their children, and the psychological resilience of the mothers explained 19% of the change in the social skill levels of the children. As a result of CAHID analyses, school type and mothers' ages on children's initial skills; fathers' education levels, mothers' ages, children's ages and genders on children's academic support skills; fathers' education levels, mothers' ages and children's ages on children's friendship skills; children's age on children's ability to manage their emotions; On the other hand, it was seen that the educational status of the mothers and the age of the children were effective on the social skills of the children. [ABSTRACT FROM AUTHOR]

After the COVID-19 epidemic, which emerged in December 2019 and is still in effect, almost all countries had to implement strict measures to control the spread of the virus. The ability of experts to reduce the spread primarily depends on the determination of the criteria affecting the spread. The fact many criteria that affect the rate of spread of COVID-19 and the most effective criteria cannot be determined, causes the spread, and therefore the number of positive cases and deaths to increase. Therefore, in the study; firstly, the weights of the criteria affecting the rate of spread were determined by using the full consistency method (FUCOM), which is a multi-criteria decision-making method, and the criteria that most affected the spread were determined by Pareto analysis, based on the criteria weights obtained. Then, based on the criteria obtained, the number of confirmed cases was predicted using the random forest method. The performance criteria values of the random forest were compared with different artificial intelligence methods such as artificial neural network, decision tree and support vector machine. Random forest gave the best results with error values (RMSE (3247), MAE (1714) and RRSE (0.374)). In addition, the random forest achieved a high prediction success of 92.9%. [ABSTRACT FROM AUTHOR]

Wind power forecasting is necessary for system operator and wind farm for voltage and frequency control, load dispatch, unit commitment, maintenance planning and electricity market actions. Wind power time series, which is a intermittent source, is not stationary and contains various outliers. This situation reduces the success of forecasting models. In order for the wind power forecasting model to give the best results, in this paper the wind power data was transformed with discrete wavelet transform. Transformed data were trained and forecested with Xgboost, a decision tree based, gradient boosting algorithm. Proposed model were designed separately for a selected month from each season. These models were compared with MAE, RMSE, R2 error metrics by the models designed with Xgboost without discrete wavelet transform. Discrete wavelet and Xgboost model gave more successful results than Xgboost model. [ABSTRACT FROM AUTHOR]

It is thought that data mining can solve many problems in visible light communication. In the VLC systems, different receiver and transmitter units are usually designed for each modulation technique. Here, the proposed systems are designed for shortened pulse position modulation. In the paper, K-Nearest Neighbor algorithm and Decision Trees based demodulators that can be applied to synchronous modulation techniques are proposed. Afterwards, a generalized entropy expression is obtained for M-SPPM signals. Additionally, demodulators that can decode signals according to the Manhattan and Minkowski distance and other demodulators are compared in terms of accuracy and bit error rate. From the numerical results, it can be seen that the proposed KNN and Decision Trees based demodulators can be used as traditional systems for synchronous modulation techniques. Furthermore, it is seen that bit error rate performances and accuracy rates of the proposed KNN based demodulator and decision trees based demodulator that is optimally designed are completely the same. In 4-SPPM signals, it is seen that the accuracy of KNN (Demodulator-1) and Decision Trees (Demodulator-3) is 98.78% for SNR=10 dB. Here, the accuracy of the KNN (Demodulator-2) and Decision Trees (Demodulator-5) is 99.07% for 8-SPPM signals. [ABSTRACT FROM AUTHOR]

Machine learning did not become widespread until the 1990s when data mining began to be used. In this study, the numerical data collected from the Central Bank of the Republic of Turkey has been divided into sections according to data groups by data mining. The algorithms used for machine learning are decision support trees, linear regression, support vector machines, gaussian regression. Exchange rate models were created by using the data for the period of January 2020 -May 2021. Four different machine learning methods were used for 1352 exchange rate data, and they were classified as 99.84% accurate in the first method, 99% in the second method, 93.72% in the third method, and 86.83% in the fourth method. It develops strategies and plans according to the highest accuracy rate Evaluations in the construction of prediction programs of data mining and machine learning methods will be useful for future applications. The results were analyzed by comparing the accuracy rates of the methods used in the literature. [ABSTRACT FROM AUTHOR]

Economic growth is accepted as the ultimate goal of all economic activities for a nation. Because of this importance, it is the first economic indicator that comes to mind both in evaluating the success of national policies and in international comparisons. Recently, various machine learning approaches and artificial neural network models have been used in this field. The aim of this study is to reveal the sources of economic growth by using decision tree modeling, different from the literature. In this context, variables accepted as sources of growth in the literature such as savings, capital accumulation, labor, innovation, public expenditures, foreign trade, human capital, demographic factors, industrialization, and institutional factors are included in the analysis. The findings of the study present rules for economic growth and provide evidence that can be a reference to policy makers for high growth rates. The findings of the study reveal the importance of savings, innovation, exports and labor in stimulating growth in the Turkish Economy. The rules obtained for economic growth provide evidence that can be a reference for policy makers to ensure high growth. [ABSTRACT FROM AUTHOR]

The mapping of water body areas such as rivers, streams, lakes and ponds is very important in terms of monitoring water resources, determining, and managing their change over time. Extracting water body areas is a complicated process that is influenced by a variety of factors. For the problem of identifying water and non-water areas, various multi spectral band satellite imagery and classification-based approaches are used. In this study, non-parametric (Support Vector Machines, k-Nearest Neighborhood and Decision Trees), probabilistic (Hidden Markov Model) and deep learning (Auto- Encoder) based supervised classification, which produce more successful results than index-based methods, were used to investigate the effectiveness of classification algorithms. Since multispectral high spatial resolution satellite images are costly, the water surface areas of Arıklar and Denizli ponds were determined by using only the red, green and blue bands of the Sentinel-2 satellite image for classification. Experimental results were compared using metrics obtained from the confusion matrix such as accuracy, specificity, precision, sensitivity, f-Score, and statistical tools such as mean square error, structural similarity index, peak signal-to-noise ratio, and Kohen's Kappa metric used in image quality determination. In the quantitative and qualitative experimental results obtained, while the deep learning-based autoencoder was the most successful method statistically, it was determined that the decision trees method worked faster in terms of time comparison. [ABSTRACT FROM AUTHOR]

Objective: In order to diagnose myeloproliferative diseases, the presence of genetic mutations is examined by the clinician in line with the disease diagnosis scheme determined by the World Health Organization. In this study, it is aimed to predict the appropriate diagnostic screening test with the model created based on bone marrow and complete blood count findings of patients who applied to the clinic. Accordingly, it is aimed to save time and financial for patients who apply to the clinic. Material and Methods: Prediction was made using the machine learning algorithms by considering the findings of patients referred to the Başkent University Ankara Hospital, Department of Medical Genetics Genetic Diseases Diagnosis Center. Descriptive statistics of the study data were given as median, and statistically significant differences were investigated using the Kruskal-Wallis test statistic. Classification algorithms such as Naive Bayes, K-Nearest Neighbor, Linear Discriminant Analysis, Support Vector Machines, Entropy Based Classification and Decision Tree were used in the estimation. With the algorithms, determinative values such as accuracy, specificity and sensitivity were obtained. The estimations made were examined according to the accuracy rates obtained and the best model was tried to be selected. Results: Statistically significant differences were found between the examined complete blood count values and the presence of mutation. The accuracy rates were found to be around 60% in the models created with Naive Bayes, K-Nearest Neighbor, Linear Discriminant Analysis, Support Vector Machines, Entropy Based Classification and Decision Tree Algorithms. Conclusion: Although the accuracy rate obtained from the machine learning algorithms used is at a moderate level, it was concluded that the results of the study would make a significant contribution to the field, since similar studies have not been included in the literature. [ABSTRACT FROM AUTHOR]

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]

Computer assisted diagnosis (CAD) systems have been used frequently in recent years in order to create a doctoral assistance decision support system using various patient information. In this study, it was aimed to compare the success of the Wechsler Intelligence Scale for Children (WISC-R) profiles by the decision trees algorithm applied to the CAD system, and the classification success in the detection of the border intelligence (BI), mild and moderate intellectual disability (ID). The data set of the study was formed by using WISC-R test results of 132 patients (50 BI, 61 mild ID and 21 moderate ID) diagnosed according to DSM-V. In order to compare the effect of WISC-R scores on the outcome, 132 patients' test scores: total, verbal and performance intelligence scores, verbal and performance intelligence subscale scores and 3 separate data sets were formed. The all the score types of the WISCR test, the first two node total intelligence scores were selected, and 108 of 132 patients were classified regardless of another attribute. In the decision tree, the first and third datasets containing the total intelligence section score type achieved a high classification success rate of 0.91. The results of this study show that the total intelligence score of the WISC-R profiles is the most effective in the decision trees algorithm in BI, mild and moderate ID diagnosis, and that CAD systems are possible. [ABSTRACT FROM AUTHOR]

Decision tree algorithm is an important classification method in data mining techniques. A decision tree creates classification and regression models like a tree that has a root node, branches, and leaf nodes. Logistic regression which is an alternative method to regression analysis when the dependent variable is a dichotomy, is another technique used for classification purposes. Within the scope of this research, logistic regression, linear regression, classification tree, and regression tree were applied on the same data set. This study explores the most important variables determining the house price by using these four methods. Models’ performances and predictive powers were compared and the best model is determined. This comparison was performed using 414 real estate data on 5 independent variables and the dependent variable is house price. The findings showed that the classification tree model for real estate valuation data performs better than standard approaches. [ABSTRACT FROM AUTHOR]

In this study, 102 businesses, 51 manipulated and 51 non-manipulated, operating in the Borsa Istanbul Manufacturing Industry have been predicted 3-9 months before whether their stock price will be manipulated. Artificial Neural Networks (ANN), Classification and Regression Tree (CART), C5.0 decision tree algorithm and Logistic Regression (LR) methods were used for classification by obtaining financial statements and qualitative data of related businesses. Overall classification accuracy was found as ANN (86.67%), CART (80.00%), C5.0 (76.67%) and LR (70.00%), respectively. The C5.0 algorithm predicted 14 of the 15 manipulated stocks and achieved 93.33% classification accuracy. ANN predicted 14 of the 15 nonmanipulated stocks and achieved 93.33% classification accuracy. In this study, since all methods have over 70% predicton performance, related models can also be included in stock manipulation prediction analyzes. [ABSTRACT FROM AUTHOR]

In our coun try, the number of small ruminant animals is decreasing day by day due to various reasons. In parallel with the decrease in the number of small ruminants, significant decreases are seen in animal production. One way to prevent the reduction in the number of small ruminants is to be able to make successful predictions and analysis related to the diagnosis. Thanks to computer-aided diagnostic studies performed with machine learning, the quality of health services increases while the costs of the health sector decrease. The aim of this study is to perform computer aided diagnosis in neonatal lambs using machine learning methods. Hence in study, decision tree, naive bayes, k-nearest neighbors, artificial neural networks and random forest methods were used. The performances of these classification methods were analyzed with accuracy, balanced accuracy, specifity, recall, F-measure, kappa and area under the ROC curve (A UC) criteria. As a result of the study, the Naive bayes method more successful results than other methods for computer aided diagnosis produced. It is very important tha t, the Naive bayes method is simple and easy to apply, achieves more successful results than other complex methods. [ABSTRACT FROM AUTHOR]

McCabe and Halstead method-level metrics are among the well-known and widely used quantitative software metrics are used to measure software quality in a concrete way. Software defect prediction can guess which or which of the sub-modules in the software to be developed may be more prone to defect. Thus, loss of labor and time can be avoided. The datasets which are used for software defect prediction, usually have an unbalanced class distribution, since the number of records with defective class can be fewer than the number of records with not defective class and this situation adversely affect the results of the machine learning methods. Information gain is employed in decision trees and decision tree based rule classifier and attribute selection methods. In this study, software metrics that provide important information for software defect prediction have been investigated and CM1, JM1, KC1 and PC1 datasets of NASA's PROMISE software repository have been balanced with the synthetic data over-sampling Smote algorithm and improved in terms of information gain. As a result, the software defect prediction datasets with higher classification success performance and the software metrics with increased information gain ratio are obtained in the decision trees. [ABSTRACT FROM AUTHOR]

The aim of the study is describing Turkish household's olive oil consumption levels and the profiles by using decision tree data mining method. The most effective factor is income level and while income level increases, the ratio of the olive oil consuming households increased. The other factors are household head's age and education level and household size. Household size is distinctive among low income households. While household size increases, the ratio of olive oil consuming households decreased. Amongst lower-middle and upper-middle income level households, while the household head's age increases, the ratio of olive oil consuming households increased. [ABSTRACT FROM AUTHOR]

As one of the ensemble methods, Mixture of Experts is used to gain higher prediction performance in classification area. In this technique, a dataset is divided into regions by soft clustering. An expert for each region is assigned and trained with the samples of the corresponding region. In this study, a dataset is divided into regions by a hard clustering method and a decision tree classifier is constructed for each region as an expert. Each expert has their individual decision for a test point. The class prediction for the test point is performed by the proposed gate function which aggregates all the decisions of experts. In calculations of the final decision via the gate function, each expert has different weights according to its distance to the test point. In the experiments on a group of benchmark datasets, better performance has been obtained with the new gate function. This new method gives better results than base classifiers such as decision trees and k Nearest Neighbors. [ABSTRACT FROM AUTHOR]

Objective: In this study, decision support system was created to diagnose jaundice and to reduce incorrect diagnosis of jaundice. The study was conducted to help physicians to diagnose jaundice. Material and Method: In this study, data of 300 patients diagnosed with jaundice was taken from a health institution. All tests performed for the patients were included in the study from the beginning of complaints to the final diagnosis. To this end, data of patients previously diagnosed with jaundice were collected in the database, edited, and organized. Training kit was created with this organized data. Then data mining methods applied to agreed algorithms and the decision tree models were created from the results. Decision trees were created by using C5.0 and J48 algorithms. The purpose of the decision tree is to reach 16 illness using 21 qualities. Results: In decision support system are two different decision trees and separate pages designed according to expert opinion. The purpose is not to create decision support system with a single decision tree, but to identify the disease with three different opinions. Trimming operations were carried out in trees created with the algorithms. With most appropriate trimming 16 illnesses were reached with 100% accuracy. Conclusion: Decision support system helping the physicians be able to use the decision trees in clinic environment easily has been developed. This study was conducted with the aim of preventing time loss in diagnosis phase and helping the physicians diagnose jaundice accurately. [ABSTRACT FROM AUTHOR]

Credit risk is one of the major risks faced by commercial banks and has an important effect on profitability ratios. Consequently, the credit scoring system in decision making for banks and other financial institutions lending to consumers is very important. Financial institutions constitute various internal credit assessment models reveal relationships between variables affecting credit scoring. In the literature, statistics and machine learning techniques for credit ratings have been widely studied. In this study, particularly for banks, financial institutions and customers of these institutions are discussed, including issues that are important to determine customer's credit score. Customer's credit score is a term used to allocate credit using statistical techniques and methods of machine learning techniques. Such methods have become increasingly more important in recent years with the growth in commercial loans. Although a wide range of applied statistical methods included in the literature, they are limited because of commercial confidentiality. In this study, for issuing the decision on the assessment of customers' demand for credit, acceptance of loan requests or refusal, a software model based on the artificial neural network (ANN) methodology is recommended. A real data set belongs to a deposit bank was used for the application. In addition to this, this study benchmarks the performance of ANN model with decision trees (DT) model. Based on the findings, ANN model outperforms the DT model in terms of estimating credit risk and customer's credit score. It is also considered that the model would be helpful in increasing the profitability of lenders gained from credits. [ABSTRACT FROM AUTHOR]

In this study, a data mining application was conducted using the data from students of Faculty of Economics and Administrative Sciences in Cumhuriyet University. Both decision tree algorithm which classifies the data best and classes produced by this algorithm were tried to determine by benefiting from the students' data. As a result of the study, it is explored that LADTree algorithm was the best algorithm which classifies the students' data and nineteen classes were produced by this algorithm. [ABSTRACT FROM AUTHOR]

The intensive increase in the competition of marketing campaigns over time reduced the impact of them on customer base. Economic pressures, intense competition in the industry, changing lifestyles of people and developing technology have caused marketing managers to adopt the concept of direct marketing by entering into new pursuits. The campaigns prepared in accordance with this understanding might be improved using a variety of data mining techniques. This study compares the performances of artifical neural networks, logistic regression and decision tree data mining techniques on a direct marketing campaign. The purpose of the study is to determine the best target group involved in the campaign by comparing estimation powers of the methods used for determining target groups. Based on the results of this study, it is revealed that artificial neural networks method is more reliable than decision tree and logistic regression analysis about estimating the likely responders in the campaign. This model can improve the efficiency of campaigns by determining of the main features that affect the success of the campaign, identifying the best target group and managing of resources. [ABSTRACT FROM AUTHOR]

Developments in computer technologies caused increase in amount of information generated and volume of database systems. By discovering data kept stored in databases as ones having beneficial use potential and creation of meaningful patterns is expressed with the concept of data mining. Decision trees are one of the data mining approaches widely used for classification and forecasting. In this study, a survey was taken on computer and internet security and it was aimed to make an inference for people have different demographic characteristics by using decision trees. [ABSTRACT FROM AUTHOR]

The aims of the current study centered on investigating some of the variables that predicted internet addiction in adolescents with the help of decision tree techniques and comparing and contrasting the differences in internet addiction prediction as a categorical and continuous variable by utilizing CART and CHAID techniques. The sample of the study consisted of 165 adolescents with normal development consulting the children's clinic of a State Hospital in Turkey for five months in 2012 for temporary problems related to the period of adolescence. It was used data collection tools which are Internet Addiction Scale, Multidimensional Scale of Perceived Social Support and personal information form. In the study, predicted or dependent variable was internet addiction whereas predictor or independent variables were 17 variables related to social relationships and internet use. In conclusion, the current study found a negative relationship between internet addiction and perceived social support from family and friends. Also, it found a positive relationship between internet addiction and the duration of internet use. The study also identified that using internet for purposes of games, chat or pornography increased the level of addiction. When internet addiction was investigated as a categorical or continuous variable, CART and CHAID analyses findings showed that clearer information and findings were obtained by including internet addiction to the analyses as a continuous variable. In addition, CART and CHAID findings were compared and it was observed that CART analysis predicted the model more correctly compared to CHAID analysis. However, it was observed that the difference was not large between CART and CHAID findings. [ABSTRACT FROM AUTHOR]

The traditional statistics method is frequently used in classification problems. However data mining techniques draw more attention day by day because they do not have any assumption and their classification performance is determined to be stronger via many researches. In this study the Decision Trees and Artificial Neural Networks data mining techniques will be introduced and the classification performances of these methods will be compared with application sample. In the application 9 macroeconomic variable data of member and candidate states of the European Union (EU) have been used. The aim of the application is to determine whether the factor of 'increasing the economical development level' is effective or not and if it is which variables are more important. The data have been analyzed via machine training package program of WEKA. C4.5 algorithm has been used for KA analysis whereas Multilayer Perceptrons have been used for YSA analysis. [ABSTRACT FROM AUTHOR]

Most of investment projects have managerial flexibility and this flexibility can significantly affect the value of projects. Because decision makers have right to alter investment decisions. But traditional valuation methods like discounted cash flow do not capture this flexibility. Flexibility can be in many forms, such as defer, expand, abandon and contract. There are two approaches that correctly capture flexibility: Decision tree analysis and real option valuation model. Real option valuation model includes managerial flexibility in terms of option and it is theoretically superior to decision tree analysis. Although decision tree analysis and real option model differ on some points, each approach has some advantages in certain applications. [ABSTRACT FROM AUTHOR]

In this study, a decision support system is designed which helps the physicians and facilitates their works for detection of various diseases from histopathological images. The developed decision support system is based on pattern recognition. The feature extraction stage of the pattern recognition system is realized by using fractal analysis and a decision tree classifier is used for classification stage. The performance of the proposed pattern recognition system is evaluated with the histopathological images which are taken from hepatic patients. The correct classification rate is 92 % for 50 patient's data. [ABSTRACT FROM AUTHOR]