A New Multi-Viewpoint and Multi-Level Clustering Paradigm for Efficient Data Mining Tasks

The database of real world contains a huge volume of data and among them there are hidden piles of interesting relations that are actually very hard to find out. The knowledge discovery in databases (KDD) appears as a possible solution to find out such relations aiming at converting information into knowledge. However, not all data presented in the bases are useful to a KDD. Usually, data are processed before being presented to a KDD aiming at reducing the amount of data and also at selecting more relevant data to be used by the system. This work consists in the use of Rough Sets Theory, in order to pre-processing data to be presented to Self-Organizing Map neural network (Hybrid Architecture) for clusters analysis. Experiments' results evidence the better performance using the Hybrid Architecture than Self-Organizing Map. The paper also presents all phases of the KDD process.

The Self Organizing Map (SOM) [1] proposed by Kohonen has proved to be remarkable in terms of its range of applications. It can be used for high dimensional space visualization, pattern recognition, input space dimensionality reduction and for generating prototyping to extrapolate information. Basically, tasks conducted by the SOM method are closely related with input space mapping in order to preserve topological and metric relationship between samples. These maps are meant to create a low dimensional output representation of high dimensional input space. Although maps higher than two dimensions can be created by SOM, it is common to work with the limit of one or two dimensions. This work presents a methodology named SOMM (Self-Organized Manifold Mapping) that can be useful to discover structures and clusters of input dataset using the SOM map as a representation of data distribution structure.KeywordsManifold LearningSelf Organizing MapsDimensionality Reduction

Segmentation and grading of brain tumors on apparent diffusion coefficient images using self-organizing maps

A Knowledge discovery in databases (KDD) system with probability deduction capability is expected to provide more information for decision making. Based on Bacchus probability logic and formal concept analysis, we propose a logic model for KDD with probability deduction. We use formal concept analysis within the semantics of probability logic to import the notion of concept into modeling of KDD. One of the most important features of a KDD system is its ability to discover previously unknown and potentially useful patterns. We formalize the definitions of previously unknown and potentially useful patterns.

Rough set (RS) theory is a mathematics tool for handling uncertain problem. It is helpful for KDD, but expensive consumption of time and unclear expression of result are the main problem in practical application. The extended concept lattice (ECL) jis a new form of concept lattice which is gotten by introducing equivalence intension into Galois concept lattice (GCL). The ECL is an efficient tool for data analysis and knowledge discovery in database (KDD). Both ECL and RS are based on equivalence class, so the relative between them exists. This paper describes the ECL first, then discusses the relation between the ECL and RS, and describes the implementation of rough set based on ECL.

The databases of real world contains a huge volume of data and among them there are hidden piles of interesting relations that are actually very hard to find out. The knowledge discovery databases (KDD) appear as a possible solution to find out such relations aiming at converting information into knowledge. However, not all data presented in the bases are useful to a KDD. Usually, data are processed before being presented to a KDD aiming at reducing the amount of data and also at selecting more relevant data to be used by the system. The purpose of this paper is to describe a validation methodology, through of a MLP neural network, to the knowledge discovery by a Hybrid Architecture composed by Rough Sets Theory used to pre-processing the data to be presented to Self-Organizing Maps neural network, which data cluster. The experimental results motivate the use of Hybrid Architecture in knowledge discovery databases.

This paper extends a previous market microstructure model, which investigated fraction dynamics of trading strategies. Our model consisted of two parts: Genetic Programming, which acted as an inference engine for trading rules, and Self-Organizing Maps (SOM), which was used for clustering the above rules into trading strategy types. However, for the purposes of the experiments of our previous work, we needed to make the assumption that SOM maps, and thus strategy types, remained the same over time. Nevertheless, this assumption could be considered as strict, and even unrealistic. In this paper, we relax this assumption. This offers a significant extension to our model, because it makes it more realistic. In addition, this extension allows us to investigate the dynamics of market behavior. We are interested in examining whether financial markets' behavior is non-stationary, because this implies that strategies from the past cannot be applied to future time periods, unless they have co-evolved with the market. The results on an empirical financial market show that its behavior constantly changes; thus, agents' strategies need to continuously adapt to the changes taking place in the market, in order to remain effective.

The Campi Flegrei caldera, a high-risk volcanic region in southern Italy, is currently experiencing an unrest phase characterized by significant ground deformation and increasing seismic activity, including events with magnitudes up to Md 4.2 recorded in September 2023. The continuous availability of seismic data provides a valuable framework for evaluating and developing novel event detection methods. However, in regions characterized by extensive natural and anthropogenic noise, the resulting low signal-to-noise ratio poses a significant challenge to seismic event detection. This limitation is sharpened when analyses are based on data from a single seismic station.To address this challenge, the present study introduces an innovative methodology designed for single-station analysis that combines the Multiscale Entropy (MSE) algorithm with Self-Organizing Maps (SOM) and the Short-Term Average/Long-Term Average (STA/LTA) technique for seismic signal detection and clustering. Linear Predictive Coding (LPC) algorithm is also employed in conjunction with the SOM map for a preliminary stage to certify the quality of the data and check for anomalies.The analysis uses continuous seismic data recorded over six months in the Pisciarelli area of the Campi Flegrei caldera, segmented into one-minute windows. Key features, including STA/LTA ratios (computed with 1s and 30s windows) and MSE values (computed over 20 time scales using a coarse-graining operation), are extracted to encode the input vectors for SOM training. The resulting 6x6 SOM map effectively clusters the seismic traces, revealing hidden patterns and distinguishing seismic events from background seismic noise. Notably, approximately 20% of the transient signals within the nodes of the seismic event cluster were identified as uncatalogued events, demonstrating the ability of the method to detect previously unrecorded activity. In addition, the map includes different clusters that highlight the influence of environmental factors, such as precipitation occurrences or volcanic fluid emissions, on the seismic waveforms.The integration of complexity-based analysis of the MSE alongside conventional STA/LTA techniques enables improved single-station event detection, even in a noisy environment, and hints at the correlation between seismic signal complexity and volcano dynamics. These results highlight the potential of advanced clustering and feature extraction techniques to refine seismic monitoring in active volcanic environments.

Large databases make computation of knowledge discovery more and more expensive, then it is proved counter-evidently that domain knowledge hidden in the database, can often guide and restrict the search for interesting knowledge, play more roles in guiding knowledge discovery in databases. While concept lattice represents knowledge on the Hass diagram with the relationships between entities and attributes, then the knowledge can be shown with hierarchical structure on the Hass diagram, thus it is properly applied to the description of association rules mining in databases. In the paper, we discuss how to use domain knowledge to guide association rules mining on concept lattice, association rules mining on which can be shown that it represents the rules more visual, vivid and concise, if using domain knowledge, we can reduce the search space, avoid blocking unexpected discoveries, so knowledge discovery can be improved effectively.

In order to improve product quality, it is necessary to summarize, analyze and resolve problems. This paper gives a concept of Problem model, which is essentially the use of Knowledge Discovery in Database (KDD) process. The Problem model derived from problem analysis and summarizing can find problems and the association rules between different problems. The model can improve the development process, improve testing design and find more problems. According to the analysis on Problem data of GUI, this paper established the Problem Model of GUI. Also the model is applied in some projects and got a nice result.

This paper presents the use of a tin-oxide sensor array and self-organized map (SOM)-based E-nose for analysis of volatile bread aroma and explores its ability to cluster bread odor data according to the freshness of bread. A low cost tin-oxide sensor array based electronic nose system has been used for the classification of state of freshness of bread. The sensor data was acquired for a period of 3 weeks, and an unsupervised self-organizing map (SOM) model was trained using this data to correlate the sensor response to classify the bread as fresh and stale. A comparative evaluation of 3 week' of bread data was carried out using the SOM. The results suggest that the system developed is able to predict the state of bread as fresh and stale up to 98% accuracy if the test bread data sets are of the same week. The classification accuracy reduces to 75–85% if test bread data sets are from different weeks. The model is also applied on three different brands of bread and similar classification results are obtained.

Identification of flow regimes in boiling flows in a vertical annulus channel with machine learning techniques

The best known and most widely utilized pattern finding algorithm in data mining applications is association rule mining (ARM). Extraction of frequent patterns is an indispensable step in ARM. Most studies in the literature have been implemented on the concept of support and confidence framework utilization. Here, we investigated an efficient and robust ARM scheme based on a self-organizing map (SOM) and an optimized genetic algorithm (OGA). A SOM is an unsupervised neural network that efficaciously produces spatially coordinated internal feature representations and detected abstractions in the input space and is the most efficient clustering technique that reveals conventional similarities in the input space by performing a topology maintaining mapping. Hence, a SOM is utilized to generate accurate clustered frequency patterns and an OGA is used to generate positive and negative association rules with multiple consequences by studying all possible patterns. Experimental analysis on various datasets has shown the robustness of our proposed ARM in comparison to traditional rule mining approaches by proving that a greater number of positive and negative association rules is generated by the proposed methodology resulting in a better performance when compared to conventional rule mining schemes.

Association rule mining is one of the most popular pattern discovery methods used in data mining. Frequent pattern extraction is an essential step in association rule mining. Most of the proposed algorithms for extracting frequent patterns are based on the downward closure lemma concept utilizing the support and confidence framework. In this chapter we investigate an alternative method for mining frequent patterns in a transactional database. Self-Organizing Map (SOM) is an unsupervised neural network that effectively creates spatially organized internal representations of the features and abstractions detected in the input space. It is one of the most popular clustering techniques, and it reveals existing similarities in the input space by performing a topology-preserving mapping. These promising properties indicate that such a clustering technique can be used to detect frequent patterns in a top-down manner as opposed to the traditional approach that employs a bottom-up lattice search. Issues that are frequently raised when using clustering technique for the purpose of finding association rules are: (i) the completeness of association rule set, (ii) the support level for the rules generated, and (iii) the confidence level for the rules generated. We present some case studies analyzing the relationships between the SOM approach and the traditional association rule framework, and propose a way to constrain the clustering technique so that the traditional support constraint can be approximated. Throughout our experiments, we have demonstrated how a clustering approach can be used for discovering frequent patterns.

There is an intersection of knowledge between two different courses in automation major, and there is a knowledge graph between several courses. This paper takes nine courses of 127 undergraduate automation majors in a university in China as the research object. This paper attempts to find out the interrelationship and correlation intensity between courses, to guide students' self-study and teachers' lesson preparation. The research study employs the quantitative method of research. K-DD (Knowledge Discovery in Database) is used in this study. Relevance rules are used in this paper. First, we use the correlation coefficient method to calculate the correlation coefficient between the courses, and then use the correlation rule method to get the relationship between the nine courses. The algorithms in the association rules used in this paper are FP growth and Apriori algorithms. The correlation coefficient between automatic control theory and electronic measurement technology is up to 0.816. when the technical state of electronic measurement is true, the possibility that the automatic control theory is true is 0.867. When the automatic control theory state is true, the probability that the electronic measurement technology is true is 0.839. The correlation between automatic control theory and electronic measurement technology is the highest. The fact shows that there is indeed a strong correlation between the two courses.