The Journal of Cognitive Systems

Cardiovascular diseases are among the most common causes of death due to their widespread prevalence. Accurate and timely diagnosis of coronary artery disease, one of the fatal cardiovascular diseases, is very important. Angiography, an invasive method, is an expensive and special method used to determine the disease and can cause serious complications. Therefore, cheaper and more efficient data mining methods are used in the diagnosis and treatment of cardiovascular diseases. As an alternative approach, by establishing clinical decision support systems using data modeling and analysis methods such as data mining, errors and costs can be reduced by providing clinicians with computer-aided diagnosis, and patient safety and clinical decision quality can be significantly increased. In this study, the data set on the open-source access website was used to classify cardiovascular disease and consists of patient records of 14 variables created by the Cleveland clinic. Also, machine learning methods (C5.0 Decision Tree, Support Vector Machine, Multilayer Perceptron, and Ensemble Learning)were used to determine the risk of coronary artery disease by deriving 1000 and 10000 data sets from the cardiology data set obtained from original 303 patient records. Performance evaluation of models is compared in terms of accuracy, specificity, and sensitivity. In trying to determine the most successful model in estimating the risk of coronary artery disease, the results are presented comparatively.

Mehmet Kivrak

Faruk Berat Akçeşme

Cemil Çolak

SAMPLE SIZE EFFECT ON CLASSIFICATION PERFORMANCE OF MACHINE LEARNING MODELS: AN APPLICATION OF CORONARY ARTERY DISEASE

机器学习模型分类性能的样本量效应：冠状动脉疾病应用研究

Objective: In this study, it is aimed to classify type 2 Diabetes Mellitus (DM), compare the estimates of the Artificial Neural Network models and determine the factors related to the disease by applying Multilayer Perceptron (MLP) and Radial Based Function (RBF) methods on the open-access dataset. Material and Methods: In this study, the data set named “Pima Indians Diabetes Database” was obtained from https://www.kaggle.com/uciml/pima-indians-diabetes-database. The dataset contains 768 records with 268 (34.9%) type 2 diabetes patients and 500 (65.1%) people without diabetes, which have 9 variables (8 inputs and 1 outcome). MLP and RBF methods, which are artificial neural network models, were used to classify type 2 DM. Factors associated with type 2 DM were estimated by using artificial neural network models. Results: The performance values obtained with MLP from the applied models were accuracy 78.1%, specificity 81.2%, AUC 0.848, sensitivity 71%, positive predictive value 61.7%, negative predictive value 86.8% and F-score 66%. In relation to RBF model, the performance metrics were accuracy obtained 76.8%, specificity 82.1%, AUC 0.813, sensitivity 66.0%, positive predictive value 64.6%, negative predictive value 83% and F-score 65.3%, respectively. When the effects of the variables in the data set examined in this study on Type 2 DM are analyzed; The three most important variables for the MLP model were obtained as Glucose, BMI, Pregnancies respectively. For RBF, it was obtained as Glucose, Skin Thickness, and Insulin. Conclusion: The findings obtained from this study showed that the models used gave successful predictions for Type 2 DM classification. Besides, unlike similar studies examining the same dataset, the significance values of the factors associated with the models created were estimated.

Emek Güldoğan

Zeynep Tunç

Ayça Acet

Performance Evaluation of Different Artificial Neural Network Models in the Classification of Type 2 Diabetes Mellitus

不同人工神经网络模型在2型糖尿病分类中的性能评估

Aim: In this study, it is aimed to classify breast cancer and identify related factors by applying deep learning method on open access to breast cancer dataset. Materials and Methods: In this study, 11 variables related to open access to breast cancer dataset of 569 patients shared by the University of Wisconsin were used. The deep learning model for classifying breast cancer was established by a 10-fold cross-validation method. The performance of the model was evaluated with accuracy, sensitivity, specificity, positive/negative predictive values, F-score, and area under the curve (AUC). Factors associated with breast cancer were estimated from the deep learning model. Results: Accuracy, specificity, AUC, sensitivity, positive predictive value, negative predictive value, and F-score values obtained from the model were 94.91%, 91.47%, 0.988, 96.90%, 95.42%, 95.14%, and 96.03%, respectively. In this study, when the effects of the variables in the dataset on breast cancer were evaluated, the three most important variables were obtained as area mean, concave points mean and symmetry mean, respectively. Conclusion: The findings of this study showed that the deep learning model provided successful predictions for the classification of breast cancer. Also, unlike similar studies examining the same dataset, the importance values of cancer-related factors were estimated with the help of the model. In the following studies, breast cancer classification performances can give more successful predictions thanks to different deep learning architectures and ensemble learning approaches.

Classification of Breast Cancer and Determination of Related Factors with Deep Learning Approach

乳腺癌分类及深度学习方法相关因素的确定

Deep learning has become very popular in recent years. Great progress has been made in the task of classifying images with the development of deep learning. This research utilized the deep learning methods in TensorFlow to classify the bird and airplane images. In the first step, a general framework for the classification of deep learning images, an image classification network namely airplane images and bird images are built. Then, the images were randomly chosen from the Caltech-UCSD Birds-200-2011 and Caltech 101 datasets. To correctly classify airplane and bird images, total of 1600 images used. The 1072 images used to train the network and the 528 images used to test built deep learning network. The training phase lasts only 20 epochs to achieve 100% accuracy on the train set. The test data were classified as 99.05% percent. Overall accuracy is 99.69%. This research has a certain importance to explore the use of cognitive systems approach in aviation safety.

Abdullah Erhan Akkaya

A DEEP LEARNING IMAGE CLASSIFICATION USING TENSORFLOW FOR OPTICAL AVIATION SYSTEMS

基于TensorFlow的深度学习图像分类在光学航空系统中的应用

İpek Balikçi Çiçek

Zeynep Küçükakçali

COMPARISON OF DIFFERENT DECISION TREE MODELS IN CLASSIFICATION OF ANGINA PECTORIS DISEASE

不同决策树模型在心绞痛疾病分类中的比较

Thyroid cancer is among the most common endocrine system malignancies and can recur. This demonstrates how critical it is to accurately determine the likelihood of recurrence. Consequently, maximizing patient treatment regimens and formulating follow-up strategies depend heavily on accurately estimating the probability of disease recurrence. Machine learning (ML) is used to find patterns and predict subtleties in large data sets that traditional analytical techniques cannot. Utilizing these insights in clinical practice will help manage patients' treatment and monitoring plans more effectively. In this study, we train Data Mining (DM) models to predict recurrence status (No Recurrence, Recurrence Present) in patients diagnosed with thyroid cancer and classify the trained models. The thyroid dataset is reduced from 16 to 11 attributes using Boruta (Random Forest) variable selection algorithm. Thyroid cancer is analyzed with the supervised learning model, the Associative classification (IC) method. The Ameva discretization algorithm was used to improve the performance of the models and to improve the data distribution. Classification Based Association Rules (CBAR) and Regularized Class Association Rules (RCAR) algorithms were used to build the prediction models. The performance of CBAR and RCAR algorithms in accurately classifying patients diagnosed with thyroid cancer was compared and found to be 95.8% and 94.5%, respectively. CBAR algorithm is more effective than RCAR algorithm in correctly classifying patients diagnosed with thyroid cancer.

Abdulvahap Pınar

CLASSIFICATION OF THYROID CANCER DIAGNOSIS STATUS USING RELATIONAL CLASSIFICATION METHOD

基于关系分类方法的甲状腺癌诊断状态分类

This paper presents a thorough investigation into the implementation of Digital Twin technology for the fault detection of Photovoltaic (PV) panels. With the increasing deployment of PV systems worldwide, it is crucial to ensure their reliable performance and early detection of faults. Digital Twin technology offers a promising approach to replicate and simulate the behavior of physical PV panels in real-time, enabling accurate fault detection and predictive maintenance. The paper explores the principles of Digital Twin, its application in the context of PV panels, and the development of an efficient fault detection framework. The proposed methodology is validated using real-world data and compared with traditional fault detection techniques, showcasing the potential of Digital Twin technology in improving the reliability and performance of PV systems.

Musa YILMAZ

Alfredo A. MARTİNEZ-MORALES

Fault Detection in Photovoltaic Panels Using Digital Twin Technology: A Comprehensive Study

基于数字孪生技术的光伏板故障检测：一项综合研究

This study describes the development of a location-based preventive information application supported by augmented reality and artificial intelligence. It can be very difficult to stay updated about the safety measures implemented for the settlement in our region while carrying out daily tasks. With this application created in this context, while obtaining preliminary information about our location, it will be incredibly simple for us to obtain information such as the level of measures taken for the location, the last inspection control dates of the existing equipment in the location, earthquake safety, etc. The application produced by combining augmented reality, artificial intelligence, and cloud computing technologies delivers data to us for this purpose. Individuals will be aware of the location as a result of the offered information, allowing them to act more safely. As a result, the developed application made it possible to reach all the data about the places in the applied locations and successful findings were obtained.

Aytaç Uğur YERDEN

Location-Based Artificial Intelligence and Augmented Reality-Assisted Preventive Information Application

基于位置的人工智能与增强现实辅助的预防信息应用

Diabetes mellitus, a chronic metabolic disease, is characterised by persistently high blood sugar levels. It is projected that by 2030, the number of individuals with diabetes in developing nations would rise from roughly 84 million to 228 million, placing a substantial strain on healthcare systems. Therefore, there is a need for different predictions that can be used in early diagnosis, follow-up and preventive medicine for this disease. In this study, a data mining algorithm, the association classification approach, is used to classify diabetes on an open source dataset. The performance metrics of the model are accuracy, balanced accuracy, sensitivity, specificity, positive predictive value, negative predictive value and F1-score values of 0.92, 0.78, 0.58, 0.98, 0.85, 0.93, 0.70 respectively. According to these results, the classification model based on association rules is highly successful in classifying diabetes melitus. In addition, as an output of the model, certain rules are proposed that can be used in early diagnosis, treatment and preventive medicine of diabetes mellitus.

Şeyma YAŞAR

Büşra Nur FINDIK

Rule-based prediction of diabetes mellitus using a classification based on association rules

基于关联规则分类的糖尿病规则预测

Gastric cancer is a type of cancer that occurs when cells in the stomach tissue grow and multiply abnormally. Gastric cancer usually starts in the inner layer of the stomach wall and can spread to other layers over time. This type of cancer is most common in people over the age of 50, but it can also occur in younger people. Symptoms of gastric cancer include indigestion and stomach pain, nausea and vomiting, loss of appetite and weight loss, bloody stools, fatigue and weakness. Although the exact cause of stomach cancer is not known, several risk factors have been identified. These risk factors include infection with the bacterium Helicobacter pylori, a family history of stomach cancer, consumption of excessively salty foods, smoking, heavy alcohol use and some genetic factors. Diabetes, on the other hand, is a hormonal disorder that regulates the body's blood sugar levels. Normally, an organ called the pancreas controls blood sugar by producing a hormone called insulin. Insulin helps glucose (sugar) enter the cells so that they can make energy. In diabetes, this regulation is disrupted, which can lead to high blood sugar and various health problems. The relationship between stomach cancer and diabetes is not yet fully understood. In this study, machine learning models (Stochastic Gradient Boosting, Bagged Classification and Regression Trees) based on proteomic data were used to predict the diabetes risk of 40 gastric cancer patients, 21 with DM and 19 with non-DM. Performance metrics for the optimal model (Stochastic Gradient Boosting) the accuracy, balanced accuracy, sensitivity, specificity, positive predictive value, negative predictive value and F1-score values are 0.86, 0.83, 0.67, 1.00, 1.00, 0.80, 0.80, respectively. According to the variable importance values obtained as a result of the model, Mucin-13 protein has a positive predictive value in predicting the diabetes risk of gastric cancer patients in the clinic.

Biomarkers for predicting diabetes in gastric cancer patients with machine learning methods based on proteomic data

基于蛋白质组数据的机器学习方法预测胃癌患者糖尿病的生物标志物

Publication Information	Publisher: Istanbul Technical University，Publishing cycle: ，Journal Type: journal，Open Access Journals: No
Basic data	Year of publication: ，Proportion of original research papers: ，Self Citation Rate:， Gold OA Rate:

The Journal of Cognitive Systems

Journal Citation Format

A journal article with 1 author

A journal article with 2 authors

A journal article with 3 authors

A journal article with 5 or more authors

Books Citation Format

Thesis Citation Format

Web sites Citation Format

Patent Citation Format

Staying up late manually editing references? ivySCI automatically matches journals and helps you generate references with a single click.

Share Submission Experience