Cluster Centroids Research Articles

A Deep Learning Approach to Clustering Visual Arts

Clustering artworks is difficult for several reasons. On the one hand, recognizing meaningful patterns based on domain knowledge and visual perception is extremely hard. On the other hand, applying traditional clustering and feature reduction techniques to the highly dimensional pixel space can be ineffective. To address these issues, in this paper we propose DELIUS: a DEep learning approach to cLustering vIsUal artS. The method uses a pre-trained convolutional network to extract features and then feeds these features into a deep embedded clustering model, where the task of mapping the input data to a latent space is jointly optimized with the task of finding a set of cluster centroids in this latent space. Quantitative and qualitative experimental results show the effectiveness of the proposed method. DELIUS can be useful for several tasks related to art analysis, in particular visual link retrieval and historical knowledge discovery in painting datasets.

Boosting k-means clustering with symbiotic organisms search for automatic clustering problems.

Kmeans clustering algorithm is an iterative unsupervised learning algorithm that tries to partition the given dataset into k pre-defined distinct non-overlapping clusters where each data point belongs to only one group. However, its performance is affected by its sensitivity to the initial cluster centroids with the possibility of convergence into local optimum and specification of cluster number as the input parameter. Recently, the hybridization of metaheuristics algorithms with the K-Means algorithm has been explored to address these problems and effectively improve the algorithm’s performance. Nonetheless, most metaheuristics algorithms require rigorous parameter tunning to achieve an optimum result. This paper proposes a hybrid clustering method that combines the well-known symbiotic organisms search algorithm with K-Means using the SOS as a global search metaheuristic for generating the optimum initial cluster centroids for the K-Means. The SOS algorithm is more of a parameter-free metaheuristic with excellent search quality that only requires initialising a single control parameter. The performance of the proposed algorithm is investigated by comparing it with the classical SOS, classical K-means and other existing hybrids clustering algorithms on eleven (11) UCI Machine Learning Repository datasets and one artificial dataset. The results from the extensive computational experimentation show improved performance of the hybrid SOSK-Means for solving automatic clustering compared to the standard K-Means, symbiotic organisms search clustering methods and other hybrid clustering approaches.

Superficial white matter bundle atlas based on hierarchical fiber clustering over probabilistic tractography data

The study of short association fibers is still an incomplete task due to their higher inter-subject variability and the smaller size of this kind of fibers in comparison to known long association bundles. However, their description is essential to understand human brain dysfunction and better characterize the human brain connectome. In this work, we present a multi-subject atlas of short association fibers, which was computed using a superficial white matter bundle identification method based on fiber clustering. To create the atlas, we used probabilistic tractography from one hundred subjects from the HCP database, aligned with non-linear registration. The method starts with an intra-subject clustering of short fibers (30-85 mm). Based on a cortical atlas, the intra-subject cluster centroids from all subjects are segmented to identify the centroids connecting each region of interest (ROI) of the atlas. To reduce computational load, the centroids from each ROI group are randomly separated into ten subgroups. Then, an inter-subject hierarchical clustering is applied to each centroid subgroup, followed by a second level of clustering to select the most-reproducible clusters across subjects for each ROI group. Finally, the clusters are labeled according to the regions that they connect, and clustered to create the final bundle atlas.The resulting atlas is composed of 525 bundles of superficial short association fibers along the whole brain, with 384 bundles connecting pairs of different ROIs and 141 bundles connecting portions of the same ROI. The reproducibility of the bundles was verified using automatic segmentation on three different tractogram databases. Results for deterministic and probabilistic tractography data show high reproducibility, especially for probabilistic tractography in HCP data. In comparison to previous work, our atlas features a higher number of bundles and greater cortical surface coverage.

Exploring Fine-Grained Cluster Structure Knowledge for Unsupervised Domain Adaptation

Unsupervised domain adaptation aims to leverage knowledge from a labeled source domain to learn an accurate model in an unlabeled target domain. However, many previous approaches propose to learn domain agnostic feature representations using a global distribution alignment objective, which does not consider the fine-grained cluster structures in the source and target domains. As such, the goal of this paper is to address two challenging problems: <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1) how to thoroughly explore fine-grained cluster structure knowledge in the source and target domains, 2) how to effectively incorporate these structure knowledge for adaptation.</i> Regarding the first point, we are motivated by structural domain similarity assumption and propose structural representation learning, which is achieved by enforcing structural consistency between the source and target domains while retaining their individual discriminative properties. Regarding the second point, we firstly devise a novel structural centroid-based label prediction method, which explicitly models structural representations to form discriminative source and target cluster centroids, and estimates the label distribution of each target sample through the cosine similarity between its corresponding target cluster centroid and all the other source cluster centroids. Then, we adopt clustering learning to incorporate these discriminative structure knowledge for adaptation by minimizing the KL divergence between the predictive target label distribution and an introduced auxiliary one. Comprehensive experiments and analyses on four benchmark datasets demonstrate the superiority of the proposed discriminative clustering framework.

Data Mining Motor Vehicle Testing Based On Vehicle Type Using The K-Means Method Case Study Binjai City Transportation Service

Motor vehicle testing is a series of activities to check the components in the vehicle. Motor vehicle testing is very important, because vehicles operated on the road have the potential to cause accidents. So if periodic inspections are not carried out, it cannot provide technical safety to vehicle users, because it is not known what components are lacking and what components must be repaired. In this study, motor vehicle test data will be calculated using the K-Means. The K-means algorithm is aclusteralgorithm non-hierarchical. Cluster analysis is a tool for grouping data based on variables or features. The purpose of k-means clustering, like other clustering methods, is to obtain clusters of data by maximizing the similarity of characteristics within the cluster and maximizing the differences between clusters.groups K-means clustering algorithm data based on the distance between the data and the centroid cluster Cluster with the number of motorized vehicle data based on the type of vehicle as many as 4 vehicles, namely, freight cars, MPU, buses, and betor. Cluster 1 there are 7 groups with 7 types of vehicles: 2 BUS and 5 betor where there is one type of vehicle (betor) that does not pass the test due to the type of damage at the time of testing motor vehicle. Cluster 1 is the type of vehicle that passes the motor vehicle test the most with the lowest level of damage; Cluster 2 there are 4 groups by type of vehicle: 4 Cars of Freight where 2 of them did not pass the test because of the type of damage during the motor vehicle test; Cluster 3 has 9 groups with the types of vehicles: 2 freight cars, 4 MPUs, and 3 BUS. 3 BUS and 1 MPU did not pass the test due to damage during the motor vehicle test. cluster is the cluster of vehicle types that do not pass the test with the most types of damage.

Centroids-guided deep multi-view K-means clustering

With the progress of deep learning used in unsupervised learning, deep approach based multi-view clustering methods have been increasingly proposed in recent years. However, in most of these methods, deep representation learning is not organically integrated into the multi-view clustering process. They either conduct deep representation learning and clustering in a separate manner, or use the pseudo cluster labels to supervise deep representation learning. In this paper, we propose a centroids-guided deep multi-view k-means clustering method, which organically incorporates deep representation learning into the multi-view k-means objective by using the cluster centroids in multi-view k-means to guide the deep learning of each view. In turn, more k-means-friendly representations are produced to further optimize the multi-view k-means objective. The cluster centroids of each view obtained under a common clustering partition not only represent the semantic information of the clusters but also imply consistency among different views. By reducing the loss between each representation and its assigned cluster centroid with respect to the network parameters of each view, the representations of different views will be more k-means-friendly toward a common partition. Experiments on several datasets demonstrate the effectiveness of our method.

Big data clustering using fractional sail fish-sparse fuzzy C-means and particle whale optimization based MapReduce framework

The process of retrieving essential information from the dataset is a significant data mining approach, which is specifically termed as data clustering. However, nature-inspired optimizations are designed in recent decades to solve optimization problems, particularly for data clustering complexities. However, the existing methods are not feasible to process with a large amount of data, as the execution time taken by the traditional approaches is larger. Hence, an efficient and optimal data clustering scheme is designed using the devised Fractional Sail Fish-Sparse Fuzzy C-Means + Particle Whale optimization (FSF-Sparse FCM + PWO) based MapReduce Framework (MRF) to process high dimensional data. Theproposed FSF-Sparse FCM is designed by the integration of Sail Fish Optimization (SFO) with fractional concept and Sparse FCM. The proposed MRF poses two functions, such as the mapper function and reducer function to perform the process of data clustering. Moreover, the proposed FSF-Sparse FCM is employed in the mapper phase to compute the cluster centroids, and thereby the intermediate data is generated. The intermediate data is tuned in the reducer phase using Particle Whale Optimization (PWO), which is the integration of Particle Swarm Optimization (PSO) and Whale optimization algorithm (WOA). Accordingly, the optimal cluster centroid is computed at the reducer phase using the objective function based on DB-Index. The proposed FSF-Sparse FM + PWO obtained the highest accuracy of 0.903 and lowest DB-Index of 39.07.

Debt Sustainability Assessment of Regional Budgets

The aim of the paper is to assess the debt sustainability of the budget system of the regions of the Northwestern Federal District (NWFD) of Russia and establish differentiated values of the budget constraints. The author uses methods of comparison and grouping, correlation and cluster analysis. The author’s methodology based on the methods of multivariate statistical analysis made it possible to include indicators that have the nature of leading indicators in the assessment, evaluate the integral indicator of debt sustainability, and determine the limit values of these indicators. Approbation of the author’s methodology for assessing the debt sustainability of the budget system on the statistics of the NWFD regions allowed dividing the regions into three clusters according to the types of debt sustainability and determining the threshold values of indicators for each cluster. The research results substantiate the need to reduce the high debt burden for 70% of the NWFD regions. The results indicate that the regions with a high level of debt sustainability include St. Petersburg, Leningrad and Kaliningrad regions. The integral indicator of debt sustainability of these regions is greater than one and the regions are well grouped according to six indicators of debt sustainability into a group that is homogeneous in terms of characteristics. The Republic of Komi, the Republic of Karelia, and the Arkhangelsk and Pskov regions have a low level of debt sustainability: the standard limiting values for the “debt-to-GRP ratio” indicator of the regions is less than 5%, and the “debt-to-revenue ratio” indicator — less than 42%, the value of all indicators of debt sustainability exceeds the values of the cluster centroids. For these regions, the author recommends pursuing a targeted budget policy with a mandatory debt reduction to the level of threshold values for the indicators of the regions of this cluster. The paper concludes that it is necessary to reduce the debt burden of most regions of the NWFD, as well as to establish differentiated values of budget constraint on public debt, taking into account the indicators of socio-economic development of these regions.

Data Clustering for Identification of Building Conditions Using Hybrid Multivariate Multinominal Distribution Soft Set (MMDS) Method

Identifying building conditions for user safety is an urgent matter, especially in earthquake-prone areas. Clustering buildings according to their conditions in the categories of danger, vulnerable, normal, and safe is important information for residents and the government to take further action. This study introduces a new method, namely hybrid multivariate multinomial distribution with the softest (MMDS) in working on the process of clustering building conditions into the most appropriate category and comparable to the condition data presented in the building data set. Research using the MMDS method is very important to map the condition of existing buildings in an area supported by available data sets. The results of the measurements carried out can provide information related to the building index and were clustered based on the index value of the condition of the building. The dataset used in this study is data on school buildings in the West Java region. There are 286 school building data with four condition parameters: foundation, concrete reinforcement, easel pole, and roof. From existing data and defined condition parameters, buildings can be classified accurately and in proportion to the facts on the ground. This study also compared the proposed method, MMDS, with the baseline method, namely Fuzzy Centroid Clustering (FCC) and Fuzzy k-means Clustering (FKC). The results show that the proposed method is superior to the baseline method with a faster processing time

An Improved K-means Clustering Algorithm Towards an Efficient Data-Driven Modeling

K-means algorithm is one of the well-known unsupervised machine learning algorithms. The algorithm typically finds out distinct non-overlapping clusters in which each point is assigned to a group. The minimum squared distance technique distributes each point to the nearest clusters or subgroups. One of the K-means algorithm’s main concerns is to find out the initial optimal centroids of clusters. It is the most challenging task to determine the optimum position of the initial clusters’ centroids at the very first iteration. This paper proposes an approach to find the optimal initial centroids efficiently to reduce the number of iterations and execution time. To analyze the effectiveness of our proposed method, we have utilized different real-world datasets to conduct experiments. We have first analyzed COVID-19 and patient datasets to show our proposed method’s efficiency. A synthetic dataset of 10M instances with 8 dimensions is also used to estimate the performance of the proposed algorithm. Experimental results show that our proposed method outperforms traditional kmeans++ and random centroids initialization methods regarding the computation time and the number of iterations.

Network Traffic Forecasting in Network Cybersecurity: Granular Computing Model

Industry 4.0, also known as the Internet of Things, is a concept that encompasses the joint applicability of operation, the Internet, and information technologies to expand the efficiency expectation of automation to include green and flexible processes and innovative products and services. Industrial network infrastructures must be modified to accommodate extra traffic from a variety of technologies in order to achieve this integration. In order to successfully implement cutting-edge wireless technologies, high-quality service (QoS) must be provided to end users. It is thus important to keep an eye on the functioning of the whole network without impacting base station throughput. Improved network performance is constantly needed, even for already-deployed cellular networks, such as the 4th generation (4G) and 3rd generation (3G). For the purpose of forecasting network traffic, an integrated model based on the long short-term memory (LSTM) model was used to combine clustering rough k-means (RKM) and fuzzy c-means (FCM). Clustering granules derived from FCM and RKM were also utilized to examine the network data for each calendar year. The novelty of our proposed model is the integration of the prediction and forecasting results obtained using existing prediction models with centroids of clusters. The WIDE backbone network’s live network traffic statistics were used to evaluate the proposed solution. The integrated model’s outcomes were assessed using a variety of statistical markers, including mean square error (MSE), root mean square error (RMSE), and standard error. The suggested technique was able to provide findings that were very accurate. The prediction error of LSTM with FCM was less on the basis of the MSE of 0.00783 and RMSE of 0.0885 at the training phase, where the prediction values of LSTM with the RKM had an MSE of 0.00564 and RMSE of 0.7511. Finally, the suggested model may substantially increase the prediction accuracy attained using FCM and RKM clustering.

A multi-view clustering framework via integrating K-means and graph-cut

Graph-cut and K-means are two classical clustering methods, which are used by most of existing clustering methods. First, a multi-view clustering framework via K-means and graph is proposed and a efficient clustering method is developed based on this framework. Our method calculates the discrete cluster assignment matrix directly and includes the auto-weighted strategy to consider the different contributions of views. Additionally, by generating the distance matrices for views with the corresponding graphs, our method fits both Gaussian and non-Gaussian distributed datasets. Moreover, the proposed algorithm is able to obtain robust results because it does not need initializing and computing cluster centroids. Extensive experiments on a synthetic dataset and four real-world datasets indicate the effectiveness of our method.

Research on precise positioning based on image recognition technology

Aiming at the problem that it is difficult to determine the picking location for the picking robot, precise positioning of picking grapes based on Mask-RCNN network and Cannylines detector is proposed in this paper. Firstly, the grape images under shading, direct sunlight and backlight conditions are collected, and grape images are marked with masks to make grape data sets. Then, based on the Mask-RCNN network training data set, the trained model is used to identify the grapes, and the box of framed grape cluster and the mask approximately equal to the size of grape cluster are obtained. Then, the AOI of main stem and centroid of grape cluster are calculated according to the obtained box and mask, the obtained AOI is equalized. Finally, the Cannylines detector based on Bilateral filtering is used to detect the line segment of main stem in the AOI, and the picking location is obtained according to the minimum distance between the centroid and the line segment. 150 images from the test set were used for the experiment, with 50 images selected from each type of light, including shading, direct sunlight and backlight. Experiments show that the accuracies of picking positioning under the three types of light conditions, are 0.900, 0.920, and 0.860, respectively. This method can provide precise picking location information for the grape picking robot.

Inter-cluster and intra-cluster joint optimization for unsupervised cross-domain person re-identification

The purpose of unsupervised cross-domain (UCD) person re-identification (re-ID) is to adapt the well pre-trained model on the labeled source domain to the unlabeled target domain, which tackles a more realistic problem. Most successful UCD methods are devoted to exploring the relationship at instance level in the target domain, but neglect the relationship at cluster level. However, person re-ID is the task of retrieving samples belonging to the same identity. Thus, exploring the cluster level relationship is critical to improve the performance of re-ID. In this paper, a novel inter-cluster and intra-cluster joint optimization (IcJO) model is proposed. Specifically, for the inter-cluster relationship, a novel centroid-based contrastive learning strategy is designed, making the sample close to its cluster centroid and far away from the heterogeneous centroids. In order to alleviate the influence of clustering noise, a cluster-based uncertainty estimation coefficient is introduced to guide the contrastive learning process. As for the intra-cluster relationship, a centroid-based maximum mean discrepancy (cMMD) loss is adopted to narrow the intra-cluster distribution. Moreover, in order to enhance the feature invariance of each identity, a centroid-based similarity loss is developed. Extensive experiments are conducted on three standard benchmarks. The results demonstrate that our IcJO model performs favorably against the state-of-the-art UCD person re-ID methods.

SVM-Assisted Adaptive Kernel Power Density Clustering Algorithm for Millimeter Wave Channels

Cluster-based channel modeling has gradually become a trend in the development of a channel model, since it is a good compromise between accuracy and complexity. However, most of the existing clustering algorithms require prior knowledge of clusters, initialization, and threshold choices. An accurate and automatic cluster identification algorithm is therefore highly desirable for channel modeling. In this article, adaptive kernel-power-density (AKPD) and support vector machine-assisted AKPD (SVM-AKPD) algorithms are proposed. First, a new distance-based metric is proposed to calculate an <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">adaptive-K</i> for each multipath component (MPC), in which the AKPD can be used in scenarios where we have a complex distribution of MPCs, especially for the cluster with small MPCs. Furthermore, the SVM is applied in clustering by the full partition of MPCs’ feature space to overcome the limitation of the AKPD, where the MPCs lying at a large distance from the cluster centroids will be clustered into surrounding clusters when the clusters are closely spaced in the AKPD. Finally, the performance of the proposed AKPD and SVM-AKPD is validated with measured and simulated channels data at millimeter waveband, respectively. Both numerical simulations and experimental validation results are provided to demonstrate the effectiveness and robustness of the proposed algorithm. The proposed algorithms enable applications in multiple-input–multiple-output (MIMO) channels with no prior knowledge about the clusters, such as number and initial locations. It also does not need to adjust cluster parameters manually and can be implemented for cluster-based channel modeling with a fairly low complexity.

Fast Summarization of Long Time Series with Graphics Processor

Summarization of a long time series often occurs in analytical applications related to decision-making, modeling, planning, and so on. Informally, summarization aims at discovering a small-sized set of typical patterns (subsequences) to briefly represent the long time series. Apparent approaches to summarization like motifs, shapelets, cluster centroids, and so on, either require training data or do not provide an analyst with information regarding the fraction of the time series that a typical subsequence found corresponds to. Recently introduced, the time series snippet concept overcomes the above-mentioned limitations. A snippet is a subsequence that is similar to many other subsequences of the time series with respect to a specially defined similarity measure based on the Euclidean distance. However, the original Snippet-Finder algorithm has cubic time complexity concerning the lengths of the time series and the snippet. In this article, we propose the PSF (Parallel Snippet-Finder) algorithm that accelerates the original snippet discovery schema with GPU and ensures acceptable performance over very long time series. As opposed to the original algorithm, PSF splits the calculation of the similarity of all the time series subsequences to a snippet into several steps, each of which is performed in parallel. Experimental evaluation over real-world time series shows that PSF outruns both the original algorithm and a straightforward parallelization.

Proposed Modification of K-Means Clustering Algorithm with Distance Calculation Based on Correlation

Clustering is a technique in data mining that groups a set of data into groups (clusters) of similar data. In general, there are two methods of clustering, namely the hierarchical method and the partition method. One of the most commonly used partition clustering methods in clustering is K-Means. The use of K-means method has been widely used in various fields with various purposes. Many research has been carried out to improve the performance of the K-Means method, for example, by modifying the method of determining the initial centroid or determining the appropriate number of clusters. In this research, the modification of the K-Means algorithm was carried out in calculating the distance by considering the correlation value between attributes. Attributes that have a high correlation value are assumed to have similar characteristics so that they determine the location of data in a particular cluster. The steps of the proposed method are: calculating the correlation value between attributes, determining the cluster centroid, calculating the distance by considering the value of correlation, and determining the data into certain clusters. The first contribution of this research is to propose a new distance calculation technique in the K-Means algorithm by considering correlation and the second contribution is to apply the proposed algorithm to a specific dataset, namely Iris dataset. In this research, the performance calculation of the modified algorithm was also carried out. From the experimental results using the Iris dataset, the proposed modification of the K-Means algorithm has fewer iterations than the original K-Means method, so that it requires less processing time. The original K-Means method requires 8 iterations, while the proposed method requires only 6 iterations. The proposed method also produces a higher accuracy rate of 89.33% than the original K-Means method, which is 82.67%.

New Cluster Analysis Method for Quantitative Dynamic Contrast-Enhanced MRI Assessing Tumor Heterogeneity Induced by a Tumor-Microenvironmental Ameliorator (E7130) Treatment to a Breast Cancer Mouse Model.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) can provide insight into tumor perfusion. However, a method that can quantitatively measure the intra-tumor distribution of tumor voxel clusters with a distinct range of Ktrans and ve values remains insufficiently explored. Two-dimensional cluster analysis may quantify the distribution of a tumor voxel subregion with a distinct range of Ktrans and ve values in human breast cancer xenografts. Prospective longitudinal study. Twenty-two female athymic nude mice with MCF-7 xenograft, treated with E7130, a tumor-microenvironmental ameliorator, or saline. 9.4 Tesla, turbo rapid acquisition with relaxation enhancement, and spoiled gradient-echo sequences. We performed two-dimensional k-means clustering to identify tumor voxel clusters with a distinct range of Ktrans and ve values on Days 0, 2, and 5 after treatment, calculated the ratio of the number of tumor voxels in each cluster to the total number of tumor voxels, and measured the normalized distances defined as the ratio of the distance between each tumor voxel and the nearest tumor margin to a tumor radius. Unpaired t-tests, Dunnett's multiple comparison tests, and Chi-squared test were used. The largest and second largest clusters constituted 44.4% and 27.5% of all tumor voxels with cluster centroid values of Ktrans at 0.040 min-1 and 0.116 min-1 , and ve at 0.131 and 0.201, respectively. At baseline (Day 0), the average normalized distances for the largest and second largest clusters were 0.33 and 0.24, respectively. E7130-treated group showed the normalized distance of the initial largest cluster decreasing to 0.25, while that of the second largest cluster increasing to 0.31. Saline-treated group showed no change. A two-dimensional cluster analysis might quantify the spatial distribution of a tumor subregion with a distinct range of Ktrans and ve values. 1 TECHNICAL EFFICACY: Stage 1.

Strengthening intrusion detection system for adversarial attacks: improved handling of imbalance classification problem

Most defence mechanisms such as a network-based intrusion detection system (NIDS) are often sub-optimal for the detection of an unseen malicious pattern. In response, a number of studies attempt to empower a machine-learning-based NIDS to improve the ability to recognize adversarial attacks. Along this line of research, the present work focuses on non-payload connections at the TCP stack level, which is generalized and applicable to different network applications. As a compliment to the recently published investigation that searches for the most informative feature space for classifying obfuscated connections, the problem of class imbalance is examined herein. In particular, a multiple-clustering-based undersampling framework is proposed to determine the set of cluster centroids that best represent the majority class, whose size is reduced to be on par with that of the minority. Initially, a pool of centroids is created using the concept of ensemble clustering that aims to obtain a collection of accurate and diverse clusterings. From that, the final set of representatives is selected from this pool. Three different objective functions are formed for this optimization driven process, thus leading to three variants of FF-Majority, FF-Minority and FF-Overall. Based on the thorough evaluation of a published dataset, four classification models and different settings, these new methods often exhibit better predictive performance than its baseline, the single-clustering undersampling counterpart and state-of-the-art techniques. Parameter analysis and implication for analyzing an extreme case are also provided as a guideline for future applications.

Early Identification of Abused Domains in TLD through Passive DNS Applying Machine Learning Techniques

DNS is vital for the proper functioning of the Internet. However, users use this structure for domain registration and abuse. These domains are used as tools for these users to carry out the most varied attacks. Thus, early detection of abused domains prevents more people from falling into scams. In this work, an approach for identifying abused domains was developed using passive DNS collected from an authoritative DNS server TLD along with the data enriched through geolocation, thus enabling a global view of the domains. Therefore, the system monitors the domain’s first seven days of life after its first DNS query, in which two behavior checks are performed, the first with three days and the second with seven days. The generated models apply the machine learning algorithm LightGBM, and because of the unbalanced data, the combination of Cluster Centroids and K-Means SMOTE techniques were used. As a result, it obtained an average AUC of 0.9673 for the three-day model and an average AUC of 0.9674 for the seven-day model. Finally, the validation of three and seven days in a test environment reached a TPR of 0.8656 and 0.8682, respectively. It was noted that the system has a satisfactory performance for the early identification of abused domains and the importance of a TLD to identify these domains.