While Very High-Resolution (VHR) imagery is favored for change detection due to its spatial detail, it presents challenges, notably intricate feature interactions and noise, complicating precise change identification. Addressing this, this paper introduces an unsupervised method for detecting building changes in Very High-Resolution (VHR) images, integrating the strengths of Principal Component Analysis (PCA) and K-Means clustering with a focus on building changes. Initially, PCA is employed to reduce data dimensionality, emphasizing the most significant variations across temporal datasets. The difference between the PCA-transformed images is computed, revealing areas of potential change. K-Means clustering then categorizes these regions based on their pixel values, labeling them as either changed or unchanged. A unique step in our approach is the building index extraction. This step refines the building detection by identifying contours in the segmented images based on their properties, such as area and perimeter emphasizing true building alterations and filtering out unrelated landscape changes. Experimental results on benchmark datasets, LEVIR-CD and CLCD, showcase the superior performance of the method, with an overall accuracy of 0. 97 and a Kappa coefficient of 0. 89. These results highlight the effectiveness of the proposed approach for building change detection in remote sensing and urban monitoring applications.

One of the most important aspects of software project management is the estimation of cost and time required for running information system. Therefore, software managers try to carry estimation based on behavior, properties, and project restrictions. Software cost estimation refers to the process of development requirement prediction of software system. Various kinds of effort estimation patterns have been presented in recent years, which are focused on intelligent techniques. This study made use of clustering approach for estimating required effort in software projects. The effort estimation is carried out through SWR (Step Wise Regression) and MLR (Multiple Linear Regressions) regression models as well as CART (Classification And Regression Tree) method. The performance of these methods is experimentally evaluated using real software projects. Moreover, clustering of projects is applied to the estimation process. As indicated by the results of this study, the combination of clustering method and algorithmic estimation techniques can improve the accuracy of estimates.

Introduction: Infertility is one of the problems that has caused a lot of psychological and worldly costs on infertile couples. Intrauterine Insemination (IUI) is one of the medically-Assisted Reproduction Techniques (ART) to help infertile couples to have a successful pregnancy. Because of unpredictable results of this technique, identifying the factors influencing the effectiveness of IUI is important. The aim of this study was to identify factors influencing the failure of IUI using data mining techniques.Methods: By utilizing K means algorithm, a descriptive technique of data mining, and Davis- Buldin index, the patients were divided into seven clusters and the features of each cluster were analyzed.Results: Increasing age, overweight, obesity, length and type of infertility in women appeared to be effective factors which were revealed by cluster analysis and investigation of the features of each cluster. Male factors including duration of infertility and spermogram type were other causes of failure in this method of infertility treatment.Discussion: By analyzing the results of clustering technique, the effective factors in the failure of IUI treatment in infertile couples were identified. The obtained results of clustering technique with the consultant of experts can be used for predicting the result of IUI treatment and helping researchers, physicians and infertile couples to choose the best treatment.

The Global Positioning System (GPS) is integral to the safety and efficiency of modern railway networks; however, its susceptibility to jamming and spoofing interference poses a significant threat to operational integrity. Conventional detection systems often rely on fully supervised models requiring extensive labeled data or specialized, costly hardware, limiting their scalability. This paper addresses this gap by proposing and evaluating an Enhanced Semi-Supervised K-Means (ESS-KMeans) algorithm designed to operate effectively with minimal labeled data. We compare its performance against a standard unsupervised K-Means algorithm using a challenging, synthetically generated dataset based on GPS signal characteristics such as latitude/longitude variation, altitude deviation, and Automatic Gain Control (AGC) levels. The proposed ESS-KMeans leverages a small labeled subset for robust centroid initialization and mutual information-based feature weighting, while also uniquely identifying and flagging ambiguous, low-confidence samples. Experimental results demonstrate that ESS-KMeans achieves perfect (1.000) accuracy on confidently classified samples, a significant improvement over standard K-Means (0.960), and improves cluster quality by over 45% (Silhouette Score). By delivering superior accuracy and providing a mechanism for uncertainty quantification with minimal supervision, this semi-supervised approach presents a scalable, cost-effective, and reliable solution for enhancing the resilience of railway systems against GPS interference.

Access control in the blockchain network is one of the challenges we face with the growth of the blockchain network. In the blockchain network, the set of financial activities of users that require a digital signature is performed, this information is stored in the blockchain server. Manually signing digitally and verifying the authenticity of transactions is a time consuming and user-friendly process and is one of the reasons why blockchain technology is not fully accepted. In this paper, a new method is proposed based on a combination of clustering and classification methods. First, the data is labeled using the clustering method and then the labeled data is used to teach the SVM algorithm to determine healthy transactions. The proposed method is a machine learning method for access control that automatically blocks blockchain transactions and detects abnormal transactions. In order to evaluate the proposed method, atrium data have been tested and analyzed. And with the help of KMEANS clustering algorithm and machine vector support method, healthy transactions are detected from suspects, which shows the ability to identify with 89% accuracy.

Data mining is a powerful new technique to help companies mining the patterns and trends in their customers' data, then to drive improved customer relationships, and it is one of well-known tools gi~): en to customer relationship management (CRM). Segmentation is the method of knowing the customers and partitioning a population of customers into smaller groups. The goal of this paper is developing a novel country segmentation methodology based on Recency (R), Frequency (F) and Monetary value (M) variables of Edible Fruits export from Islamic Republic of Iran to other countries during 11 years (1995-2005). After the variables are calculated, clustering methods (K-Means and fuzzy K-Means) are used to segment countries and compare the results of these methods by three different criteria. By using customer pyramid and decision tree, clusters are classified into four tiers: Loyal customer, Active customer, New customer and Inactive customer. Consequently, the data are used to analyze the relative profitability of each customer cluster and the proper CRM strategy is determined for them.

With the advancement of technology, the use of ATM and credit cards are increased. Cyber fraud and theft are the kinds of threat which result in using these Technologies. It is therefore inevitable to use fraud detection algorithms to prevent fraudulent use of bank cards. Credit card fraud can be thought of as a form of identity theft that consists of an unauthorized access to another person's card information for the purpose of charging purchases to the account or removing funds from it. Credit card fraud schemes are divided into two categories: application fraud and account takeover. When a credit card account gets opened without someone’s permission is called application fraud. Account takeovers, on the other hand, is when an existing credit card account is hijacked, and the criminal obtains enough personal information to modify the account's information. The criminal then subsequently reports the card lost or stolen in order to obtain a new card and make unauthorized purchases with it. Data mining as a technique capable of identifying useful patterns among a great deal of data is an effective method in detecting fraud in this regard. The main purpose of this paper is to present a new method for detecting unattended outliers that require high accuracy and recall. The method presented in this study is based on a combination of NMF, hierarchical K-Means, K-Means and k-nearest neighbors’ techniques. To evaluate the proposed method of outlier detection, several experiments were performed using standard data, in terms of accuracy and recall with Isolation Forest, k-nearest neighbors, Median kNN, and Average kNN. The dataset used in this paper is one that was provided in a 2016 Kaggle competition and was provided by a European bank after anonymization. The results, corroborate that the proposed method has higher accuracy and recall than other algorithms.