The objective of this research is marine Data users Clustering using Data mining technique. To achieve this objective, marine organizations will enable to know their Data and users requirements. In this research, CRISP-DM standard model was used to implement the Data mining technique. The required Data was extracted from 500 marine Data users profile Database of Iranian National Institute for Oceanography and Atmospheric Sciences (INIOAS) from 1386 to 1393. The TwoStep algorithm was used for Clustering. In this research, patterns was discovered between marine Data users such as student, organization and scientist and their Data request (Data source, Data type, Data set, Parameter and Geographic area) using Clustering for the first time. The most important clusters are: Student with International Data source, Chemistry Data type, “World Ocean Database” Dataset, Persian Gulf geographic area and Organization with Nitrate parameter. Senior managers of the marine organizations will enable to make correct decisions concerning their existing Data. They will direct to planning for better Data collection in the future. Also Data users will guide with respect to their requests. Finally, the valuable suggestions were offered to improve the performance of marine organizations.

Clustering is a widespread Data analysis and Data mining technique in many fields of study such as engineering, medicine, biology and the like. The aim of Clustering is to collect Data points. In this paper, a Cultural Algorithm (CA) is presented to optimize partition with N objects into K clusters. The CA is one of the effective methods for searching into the problem space in order to find a near optimal solution. This algorithm has been tested on different scale Datasets and has been compared with other well-known algorithms in Clustering, such as K-means, Genetic Algorithm (GA), Simulated Annealing (SA), Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO) algorithm. The results illustrate that the proposed algorithm has a good proficiency in obtaining the desired results.

Clustering has been one of the main building blocks in the fields of machine learning and computer vision. Given a pair-wise distance measure، it is challenging to find a proper way to identify a subset of representative exemplars and its associated cluster structures. Recent trend on big Data analysis poses a more demanding requirement on new Clustering algorithm to be both scalable and accurate. A recent advance in graph-based Clustering extends its ability to millions of Data points by massive utility of engineering endeavor and parallel optimization. However، most other existing Clustering algorithms، though promising in theory، are limited in the scalability issue. In this paper، a novel Clustering method is proposed that is both accurate and scalable. Based on a simple criteria، ” key” items that are representative of the whole Data set are iteratively selected and thus form associated cluster structures. Taking input of pairwise distance measure between Data instances، the proposed method searches centers of clusters by identifying Data items far away from selected keys، but representative of unselected Data items. Inspired by hierarchical Clustering، small clusters are iteratively merged until a desired number of clusters are obtained. To solve the scalability problem، a novel tracking table technique is designed to reduce the time complexity which is capable of Clustering millions of Data points within a few minutes. To assess the performance of the proposed method، several experiments are conducted. The first experiment tests the ability of our algorithm on different manifold structures and various number of clusters. It is observed that our Clustering algorithm outperforms existing alternatives in capturing different shapes of Data distributions. In the second experiment، the scalability of our algorithm to large scale Data points is assessed by Clustering up to one million Data points with dimensions of up to 100. It is shown that، even with one million Data points، the proposed method only takes a few minutes to perform Clustering. The third experiment is conducted on the ORL Database، which consists of 400 face images of 40 individuals. The proposed Clustering method outperforms the compared alternatives in this experiment as well. In the final experiment، shape Clustering is performed on the MPEG-7 Dataset، which contains 1400 silhouette images from 70 classes، 20 different shapes for each class. The goal here is to cluster the Data items (here the binary shapes) into 70 clusters، so that each cluster only includes shapes that belong to one class. The proposed method outperforms other alternative Clustering algorithms on this Dataset as well. Extensive empirical experiments demonstrate the superiority of the proposed method over existing alternatives، in terms of both effectiveness and efficiency. Furthermore، our algorithm is capable of large-scale Data Clustering where millions of Data points can be clustered in a few seconds.

Distance-based Clustering methods categorize samples by optimizing a global criterion, finding ellipsoid clusters with roughly equal sizes. In contrast, density-based Clustering techniques form clusters with arbitrary shapes and sizes by optimizing a local criterion. Most of these methods have several hyper-parameters, and their performance is highly dependent on the hyper-parameter setup. Recently, a Gaussian Density Distance (GDD) approach was proposed to optimize local criteria in terms of distance and density properties of samples. GDD can find clusters with different shapes and sizes without any free parameters. However, it may fail to discover the appropriate clusters due to the interfering of clustered samples in estimating the density and distance properties of remaining unclustered samples. Here, we introduce Adaptive GDD (AGDD), which eliminates the inappropriate effect of clustered samples by adaptively updating the parameters during Clustering. It is stable and can identify clusters with various shapes, sizes, and densities without adding extra parameters. The distance metrics calculating the dissimilarity between samples can affect the Clustering performance. The effect of different distance measurements is also analyzed on the method. The experimental results conducted on several well-known Datasets show the effectiveness of the proposed AGDD method compared to the other well-known Clustering methods.

In the last few decades, in many research fields, different methods were introduced to discover groups with the same trends in longitudinal Data. The Clustering process is an unsupervised learning method, which classifies longitudinal Data based on different criteria by performing algorithms. The current study was performed with the aim of reviewing various methods of longitudinal Data Clustering, including two general categories of non-parametric methods and model-based methods. PubMed, SCOPUS, ISI, Ovid, and Google Scholar were searched between 2000 and 2021. According to our systematic review, the non-parametric k-means Clustering Method utilizing Euclidean distance emerges as a leading approach for Clustering longitudinal Data This research, with an overview of the studies done in the field of Clustering, can help researchers as a toolbox to choose various methods of longitudinal Data Clustering in idea generation and choosing the appropriate method in the classification and analysis of longitudinal Data.

Data mining and knowledge discovery are important technologies for business and research. Despite their benefits in various areas such as marketing, business and medical analysis, the use of Data mining techniques can also result in new threats to privacy and information security. Therefore, a new class of Data mining methods called privacy preserving Data mining (PPDM) has been developed. The aim of researches in this field is to develop techniques those could be applied to Databases without violating the privacy of individuals. In this work we introduce a new approach to preserve sensitive information in Databases with both numerical and categorical attributes using fuzzy logic. We map a Database into a new one that conceals private information while preserving mining benefits. In our proposed method, we use fuzzy membership functions (MFs) such as Gaussian, P-shaped, Sigmoid, S-shaped and Z-shaped for private Data. Then we cluster modified Datasets by Expectation Maximization (EM) algorithm. Our experimental results show that using fuzzy logic for preserving Data privacy guarantees valid Data Clustering results while protecting sensitive information. The accuracy of the Clustering algorithm using fuzzy Data is approximately equivalent to original Data and is better than the state of the art methods in this field.

In the real world Clustering problems, it is often encountered to perform cluster analysis on Data sets with mixed numeric and categorical values. However, most existing Clustering algorithms are only efficient for the numeric Data rather than the mixed Data set. In addition, traditional methods, for example, the K-means algorithm, usually ask the user to provide the number of clusters. In this paper, we propose a new method to cluster mixed Data and automatically evolve the number of clusters as well as Clustering of Data set. In the proposed method, Davies-Bouldin Index is used as fitness function and we use the genetic algorithm to optimize fitness function. Also, we use a more accurate distance measure for calculating the distance between categorical values. The performance of this algorithm has been studied on real world and simulated Data sets. Comparisons with other Clustering algorithms illustrate the effectiveness of this approach.