Clustering is a widely utilized data mining technique aimed at discovering unknown patterns in datasets, typically performed in an unsupervised manner. However, the incorporation of external knowledge, transforming it into a semi-supervised process, has shown promising results in improving clustering performance. This study explores the application of feature constraints in clustering algorithms, specifically focusing on relative constraints, which are effective in embedding domain knowledge. We introduce a method to use relative constraints as features, integrating them into the Neighborhood-Based Clustering algorithm. Our experiments demonstrate the impact of these constraints on clustering accuracy and efficiency using various benchmark datasets. The results indicate that the proposed method enhances clustering quality by effectively incorporating domain-specific constraints, offering a novel approach to semi-supervised clustering.

This paper proposes a new approach to the k -means clustering algorithm incorporating a variety of fuzzy equivalences and aggregation techniques. Central to this approach is the use of fuzzy equivalences, which serve as an alternative to standard distance metrics, with the goal of improving the clustering process. The modified k-means algorithm employs these fuzzy equivalences to better determine the similarity between the data points, particularly in situations where the nearest points may not adequately represent closeness within the data set. To assess the clustering results, we employ a variation of the silhouette coefficient tailored to our method. Additionally, we present theoretical insights into the behavior and benefits of using composition of aggregations and fuzzy equivalences in clustering. Experimental validations carried out on diverse data sets indicate that our method can lead to improved clustering outcomes compared to traditional techniques.

Data clustering algorithms are vital in artificial intelligence and data mining to uncover unknown patterns in data sets. This paper investigates the use of data clustering and visualization techniques to analyze medical data sets, with a focus on the application of fuzzy equivalence in clustering. Fuzzy equivalences provide a substitute for conventional distance metrics, improving the clustering procedure by allowing the selection of the optimal mix of fuzzy distance and aggregation for the data set under analysis. We explain how these equivalences are incorporated into clustering algorithms and their impact on visualizing medical data. Our approach determines the most suitable fuzzy equivalence and aggregation methods for data sets, based on the silhouette coefficient, and explores visualization scenarios to identify distinctive data set attributes. Hence, the incorporation of fuzzy equivalences and aggregations into clustering algorithms enables more adaptable tuning of clustering results based on the silhouette coefficient, and the suggested visualization facilitates a quick visual assessment of important data set characteristics.

In this paper, we examine the possibility of employing the idea of progressive-inductive (π) aggregation in the k-means algorithm. We base our work on the interactive visualization framework called Skydive which is a tightly coupled system that combines data aggregation and visual aggregation for smooth user interaction. It is based on an aggregate pyramid data structure (π-cube), allowing a visualization client to quickly filter the data. Skydive was designed for interactive data visualization in terms of expressiveness and efficiency. Therefore, the underlying data structure enabled access to data on desired levels of granularity. In this paper, we utilize this concept to speed up the process of clustering proposing π-means - a modified version of the k-means algorithm. This will allow us to explore further possible applications of π-cube towards interactive data exploration.

For modern manufacturing firms, automation has already become a norm but constantly needs to be improved as firms still face strong demand to increase their productivity. This can be achieved by reducing dependability on manpower, reaching lean and even unmanned production and this is where some of the standards of Industry 4.0 come in useful, not to mention: Machine Vision, Image Recognition or Machine Learning. In our paper, we present SODA – our approach to build a flexible ML and AI enabled framework for object detection, analysis, and simulation. The framework is designed to support a development process of solutions requiring real-time analysis of images of different types of moving objects on a conveyor belt. In our work we discuss architectural challenges of the developed framework as well as the basic components of the system. We do also provide information on how to use the framework and present a sample implementation of an actual system employing some of the machine learning methods.

Data clustering algorithms have proved to be important and widely used methods of artificial intelligence and data mining for discovering unknown yet important patterns in datasets. Nevertheless, one of the additional aspects of data clustering is proper interpretation of the clustering results. In this paper we aim to investigate possibilities of using both data clustering and visualization methods to analyze a sample diabetes dataset. In the first part, we focus on how to cluster a highly-dimensional sample dataset and then, we concentrate on how to properly visually present the clustering results in the most meaningful way to uncover potentially interesting behavioral patterns or features of diabetes patients. In this work we examine two clustering algorithms (DBSCAN, k-Means) along with several different distance measures. We also present sample visualizations of clustering results generated by an application which we have developed and discuss if the proposed way of clustering results visualization can be helpful in understanding the analyzed dataset and lead a viewer to drawing valuable conclusions about it.

In 1973 Lotfi Zadeh introduced the theory of fuzzy logic [17]. Fuzzy logic was an extension of Boolean logic so that it allowed using not only Boolean values to express reality. One kind of basic logical operations in fuzzy logic are so-called fuzzy implications. From over eight decades a number of different fuzzy implications have been described [3] - [16]. In the family of all fuzzy implications the partial order induced from [0,1] interval exists. Pairs of incomparable fuzzy implications can generate new fuzzy implications by using min(inf) and max(sup) operations. As a result the structure of lattice is created ([1], page 186). This leads to the following question: how to choose the correct functions among basic fuzzy implications and other generated as described above. In our paper, we propose a new method for choosing implications. Our method allows to compare two fuzzy implications. If the truth value of the antecedent and the truth value of the implication are given, by means of inverse fuzzy implications we can easily optimize the truth value of the implication consequent. In other words, we can choose the fuzzy implication, which has the greatest or the smallest truth value of the implication consequent or which has greater or smaller truth value than another implication. Primary results regarding this problem are included in the paper [14].

Visualization provides a powerful means for data analysis. But to be practical, visual analytics tools must support smooth and flexible use of visualizations at a fast rate. This becomes increasingly onerous with the ever-increasing size of real-world datasets. First, large databases make interaction more difficult once query response time exceeds several seconds. Second, any attempt to show all data points will overload the visualization, resulting in chaos that will only confuse the user. Over the last few years, substantial effort has been put into addressing both of these issues and many innovative solutions have been proposed. Indeed,data visualizationis a topic that is too large to be addressed in a single survey paper. Thus, we restrict our attention here tointeractivevisualization oflargedata sets. Our focus then is skewed in a natural way towards query processing problem—provided by an underlying database system—rather than to the actual data visualization problem.