Balanced Iterative Reducing And Clustering Using Hierarchies Research Articles

An operation health status monitoring algorithm of special transformers based on BIRCH and Gaussian cloud methods

The health status monitoring of special transformers is of great significance to ensure the secure operation of the distribution network and the power quality of special transformer users. Given this background, an operation health status monitoring algorithm of special transformers is proposed in this paper based on balanced iterative reducing and clustering using hierarchies (BIRCH) and Gaussian Cloud methods (GCM). The algorithm is composed of two parts, i.e., the offline and online parts. For the offline part, the operating indexes of special transformers are extracted based on historical operating data, and Gaussian clouds of normal operating conditions of the special transformers are determined by BIRCH clustering and Gaussian cloud methods. For the online part, Gaussian clouds of real-time operating conditions of special transformers are determined by BIRCH clustering and Gaussian cloud methods based on real-time operation data. Then, the monitoring results of operating health status are determined by the distance between the standard Gaussian clouds and the real-time Gaussian clouds of special transformers. Finally, case studies for actual special transformers are performed to verify the proposed method, and the results show that the proposed model can effectively identify the abnormal operation of the special transformer.

Detection of non-typical users of the electronic marketplace "Freight transportation" to prevent the competitive intelligence

This article presents results of usage of cluster analysis and machine learning to detect non-typical users of the electronic freight transportation marketplace. The most significant red flags to identify system users who collect commercial information have been identified, a competitor’s profile has been drawn up. The use of various methods of clustering and classification in the modeling of this problem is studied. The procedure of formation of a training sample in the absence of labeled data has been carried out. The algorithm has been developed to detect non-typical users based on machine data analysis methods in order to prevent the real-time commercial intelligence. This algorithm is based on density-based spatial clustering of applications with noise (DBSCAN) and balanced iterative reducing and clustering using hierarchies (BIRCH) cluster analysis methods, identification of anomalous objects using the Isolation Forest method and ensemble Gradient Boosting algorithm with decision trees as the base classifiers. It can be used by electronic marketplace’s owners, as well as by other companies that apply similar technologies while providing their services. As a result, software tool in Python has been developed, which allows to timely detect users collecting commercial information and hence to reduce a negative impact from possible competitive intelligence.

Novel electricity pattern identification system based on improved I-nice algorithm

As a result of the rapid increase in smart-metre popularity, a large volume of smart electricity data has been generated. The underlying information contained within such data sets is very helpful and useful to businesses and power companies. However, effective mining of this valuable knowledge is challenging for both industry and academia. Current research mainly focuses on application of statistical analyses to available smart electricity data rather than construction of an advanced learning system. This paper reports development of an intelligent system to identify electricity patterns within industrial electricity data. The core component of this novel identification system is the I-nice-α clustering algorithm, which is a variant of the I-nice algorithm. I-nice-α utilises kernel density estimation technology and the minimum mean discrepancy to optimise the cluster number and determine the cluster centres, respectively, of the target data. Our theoretical analysis proves that I-nice-α has lower computational complexity than I-nice. We also compare I-nice-α with I-nice and four other standard clustering algorithms (i.e., the Gaussian mixture model (GMM), density-based spatial clustering of applications with noise (DBSCAN), the fuzzy c-means (FCM) algorithm, and balanced iterative reducing and clustering using hierarchies (BIRCH)), based on 15 benchmark data sets and electricity consumption data. The experiment results show that I-nice-α achieves superior clustering performance; thus, the feasibility and effectiveness of the improved algorithm are demonstrated. In addition, hierarchical (i.e., daily and annual) electricity patterns are determined by analysing industrial electricity data with I-nice-α; this result elucidates the production organism and power assignment. Thus, the proposed electricity pattern identification system has considerable potential application among power companies and businesses.

Speech Personality Recognition Based on Annotation Classification Using Log-Likelihood Distance and Extraction of Essential Audio Features

Speech personality recognition relies on training models that require an excessive number of features and are, in most cases, designed specifically for certain databases. As a result, when tested on different datasets, overfitted classifier models are not always reliable because their accuracy changes with changes in the domain of speakers. Moreover, personality annotations are often subjective, which creates variability in raters perception during labeling. These problems inhibit the effectiveness of speech personality recognition applications. To reduce the unexplained variance caused by unknown differences in raters perception, a structure that uses Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH) algorithm is proposed. Furthermore, a feature extraction method is proposed to filter out undesirable adulterations be it noise, silence, or uncertain pitch segments, while extracting essential audio features, i.e., signal power roll-off, pitch, and pause rate. Experiments on the standard SSPNet dataset records a relative 4% increase in overall accuracy when log-likelihood based annotations are used. Moreover, improved consistency in accuracy is observed when this method is tested on male and female subsets.

Active contour modal based on density-oriented BIRCH clustering method for medical image segmentation

Currently, medical image segmentation has attracted more attention from researchers, which can assist in medical diagnosis. However, in the process of traditional medical image segmentation, it is sensitive to the initial contour and noise, which is difficult to deal with the weak edge image, complex iterative process. In this paper, we propose a new medical image segmentation method, which adopts density-oriented BIRCH (balanced iterative reducing and clustering using hierarchies) clustering method to modify active contour model and improve the robustness of noise. The BIRCH is a multi-stage clustering method using clustering feature tree. The improved model can effectively deal with the gray non-uniformity of real medical images. And we also introduce a new energy function in active contour model to make the contour curve approach to the edge, and finally stay at the edge of the image to complete the object segmentation. Experimental results show that this new model can overcome the influence of complex background on medical image segmentation and improve the speed and accuracy of medical segmentation results.

Data-Driven Analytics for Personalized Medical Decision Making

The study was conducted by applying machine learning and data mining methods to treatment personalization. This allows individual patient characteristics to be investigated. The personalization method was built on the clustering method and associative rules. It was suggested to determine the average distance between instances in order to find the optimal performance metrics. The formalization of the medical data preprocessing stage was proposed in order to find personalized solutions based on current standards and pharmaceutical protocols. The patient data model was built using time-dependent and time-independent parameters. Personalized treatment is usually based on the decision tree method. This approach requires significant computation time and cannot be parallelized. Therefore, it was proposed to group people by conditions and to determine deviations of parameters from the normative parameters of the group, as well as the average parameters. The novelty of the paper is the new clustering method, which was built from an ensemble of cluster algorithms, and the usage of the new distance measure with Hopkins metrics, which were 0.13 less than for the k-means method. The Dunn index was 0.03 higher than for the BIRCH (balanced iterative reducing and clustering using hierarchies) algorithm. The next stage was the mining of associative rules provided separately for each cluster. This allows a personalized approach to treatment to be created for each patient based on long-term monitoring. The correctness level of the proposed medical decisions is 86%, which was approved by experts.

Partitioned Relief-F Method for Dimensionality Reduction of Hyperspectral Images

The classification of hyperspectral remote sensing images is difficult due to the curse of dimensionality. Therefore, it is necessary to find an effective way to reduce the dimensions of such images. The Relief-F method has been introduced for supervising dimensionality reduction, but the band subset obtained by this method has a large number of continuous bands, resulting in a reduction in the classification accuracy. In this paper, an improved method—called Partitioned Relief-F—is presented to mitigate the influence of continuous bands on classification accuracy while retaining important information. Firstly, the importance scores of each band are obtained using the original Relief-F method. Secondly, the whole band interval is divided in an orderly manner, using a partitioning strategy according to the correlation between the bands. Finally, the band with the highest importance score is selected in each sub-interval. To verify the effectiveness of the proposed Partitioned Relief-F method, a classification experiment is performed on three publicly available data sets. The dimensionality reduction methods Principal Component Analysis (PCA) and original Relief-F are selected for comparison. Furthermore, K-Means and Balanced Iterative Reducing and Clustering Using Hierarchies (BIRCH) are selected for comparison in terms of partitioning strategy. This paper mainly measures the effectiveness of each method indirectly, using the overall accuracy of the final classification. The experimental results indicate that the addition of the proposed partitioning strategy increases the overall accuracy of the three data sets by 1.55%, 3.14%, and 0.83%, respectively. In general, the proposed Partitioned Relief-F method can achieve significantly superior dimensionality reduction effects.

Semi-Supervised Classification of Data Streams by BIRCH Ensemble and Local Structure Mapping

Many researchers have applied clustering to handle semi-supervised classification of data streams with concept drifts. However, the generalization ability for each specific concept cannot be steadily improved, and the concept drift detection method without considering the local structural information of data cannot accurately detect concept drifts. This paper proposes to solve these problems by BIRCH (Balanced Iterative Reducing and Clustering Using Hierarchies) ensemble and local structure mapping. The local structure mapping strategy is utilized to compute local similarity around each sample and combined with semi-supervised Bayesian method to perform concept detection. If a recurrent concept is detected, a historical BIRCH ensemble classifier is selected to be incrementally updated; otherwise a new BIRCH ensemble classifier is constructed and added into the classifier pool. The extensive experiments on several synthetic and real datasets demonstrate the advantage of the proposed algorithm.

On solving large p-median problems

Incorporating big data in urban planning has great potential for better modeling of urban dynamics and more efficiently allocating limited resources. However, big data may present new challenges for problem solutions. This research focuses on the p-median problem, one of the most widely used location models in urban and regional planning. Similar to many other location models, the p-median problem is non-deterministic polynomial-time hard (NP-hard), and solving large-sized p-median problems is difficult. This research proposes a high performance computing-based algorithm, random sampling and spatial voting, to solve large-sized p-median problems. Instead of solving a large p-median problem directly, a random sampling scheme is introduced to create smaller sub- p-median problems that can be solved in parallel efficiently. A spatial voting strategy is designed to evaluate the candidate facility sites for inclusion in obtaining the final problem solution. Tests with the Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH) data set show that random sampling and spatial voting provides high-quality solutions and reduces computing time significantly. Tests also demonstrate the dynamic scalability of the algorithm; it can start with a small amount of computing resources and scale up and down flexibly depending on the availability of the computing resources.

Grouping of Medicinal Drugs Used for Similar Symptoms by Mining Clusters from Drug Benefits Reviews

Clustering is an unsupervised process for partitioning a set of data into a set of meaningful sub-classes or sub-datasets, called clusters. In this paper, the k-means clustering algorithm is used on a textual dataset of unlabeled reviews of medicinal drugs in order to group the drugs with similar usage and benefits. Two alternative clustering algorithms, agglomerative clustering and BIRCH (balanced iterative reducing and clustering using hierarchies), are used on the same dataset for comparison. The methodology for text preprocessing and feature extraction is explained. The results are analysed using quantitative performance metrics (Silhouette Score and Calinski-Harabasz Score), cluster sizes, as well as visualization by the means of scatter plot of clusters, distances from final centroids, word clouds and dendrogram.

Prediction of geological conditions for a tunnel boring machine using big operational data

This paper presents a comprehensive procedure to predict geological conditions (i.e., rock mass types) for a tunneling boring machine (TBM) based on big operational data including four channels: cutterhead speed, cutterhead torque, thrust, and advance rate. To handle the big operational data, a Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH) algorithm is adopted to effectively compress 12,038,636 TBM operational data to only 5014 leaf node entries. A K-means++ algorithm is used to find potential rock mass types in the TBM operational data. By comparing three kinds of classifiers, a support vector classifier (SVC) with an average precision of 98.6% is selected as the best geological conditions prediction model. Test results on historical TBM operational segments show that most adjacent operational segments have the same rock mass type. The change in rock mass type is a dynamic process, which first fluctuates between two rock mass types and gradually stabilizes at the latter type. In addition, the cutterhead torque and thrust are found to better reflect the change of rock mass types compared with the advance rate and cutterhead speed. Test results on a water conveyance tunnel show that using only 20% of TBM training data the developed prediction model can generate 84.4% precision and 88.8% recall performance for the remaining 80% testing data. Hence, the proposed procedure could be applied to big TBM operational data to accurately detect, characterize, and predict rock mass types, which is of critical importance to safe and efficient tunneling.

Maximum delay anonymous clustering feature tree based privacy-preserving data publishing in social networks

Clustering analysis has been widely used in pattern recognition and image processing in recent years, which is an important research field of data mining. Data publishing in social networks is threatened by the leakage of private information nowadays. This paper proposes a privacy preservation scheme of sensitive data publishing in social networks based on Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH) algorithm to tackle this issue. The scheme is divided into an online process and an offline process. Specifically, we present the Maximum Delay Anonymous Clustering Feature (MDACF) tree data publishing algorithm.

A data cleaning method for water quality based on improved hierarchical clustering algorithm

Identifying and rectifying incomplete water quality data is of vital importance. A data cleaning method based on improved balanced iterative reducing and clustering using hierarchies (BIRCH) clustering algorithm is proposed. The clustering feature tree of water quality data is constructed and the cluster vector of the clustering feature tree is obtained by the agglomerative method. The optimal cluster number is determined according to the Bayesian Information Criterion and the nearest clustering ratio. The Pauta criterion is used to detect the global outlier and artificial neural network (ANN) is used to fill in outliers and missing values. Finally, the improved data cleaning method is applied to water quality monitoring data of Beijing wastewater treatment plant. The experimental results show that the data cleaning method can not only detect abnormal values and missing values accurately, but also normalise and complete missing data.

Modified balanced iterative reducing and clustering using hierarchies (m-BIRCH) for visual clustering

In modern visual clustering applications where datasets are large and updates with new data may be ongoing, methods of online clustering are extremely important. Online clustering algorithms incrementally cluster the data points, use a fraction of the dataset memory, and update the clustering decisions when new data comes in. In this paper we adapt a classic online clustering algorithm called balanced iterative reducing and clustering using hierarchies (BIRCH) to incrementally cluster large datasets of features commonly used in visual clustering, e.g., 840 K color SIFT descriptors, 1.09 million color patches, 60 K outlier corrupted grayscale patches, and 700 K grayscale SIFT descriptors. We use the algorithm to cluster datasets consisting of non-convex clusters, e.g., Hopkins 155 3D motion segmentation dataset. We call the adapted version modified-BIRCH (m-BIRCH). BIRCH was originally developed by the database management community, but has not been used in computer vision. Modifications made in m-BIRCH enable data-driven parameter selection and effectively handle varying density regions in the feature space. Data-driven parameter selection automatically controls the level of coarseness of the data summarization. Effective handling of varying density regions is necessary to well represent the different density regions in the data summarization. Our implementation of the algorithm provides a useful clustering tool and is made publicly available.

Advanced Split BIRCH Algorithm in Reconfigurable Network

The Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH) has a disadvantage that it reduced the accuracy of the arbitrary shape clustering algorithm clusters, to this end a split improved BIRCH algorithm (AS-Birch) was put forward. Through the analysis of the reconfigurable network and a detailed analysis of application scenarios and functional requirements of business clustering, according to the different characteristics of each clustering algorithm, we choose a hierarchical clustering algorithm BIRCH as a business solution to the clustering programs. On the basis of the drawbacks in BIRCH algorithm, we designed split improved BIRCH algorithm and the specific steps of the algorithm is given. Finally, Advanced Split BIRCH algorithm was applied to a detailed experimental simulation experiments show in reconfigurable network. The experiment suggests: AS-BIRCH algorithm has a greater improvement in terms of packet loss rate, delay and jitter than BIRCH algorithm, especially when dealing with irregular shape cluster, AS-BIRCH algorithm greatly improved the accuracy rate concerning to the business clustering.

Anomaly detection model based on support vector machine and Bayesian classification

Through the research into the types of network attack and the intrusion detection methods,the fact that the normal intrusion detection method was not good enough for detecting U2R(User to Root) and R2L(Remote to Local) was found.To improve the detection rate of anomaly detection system for U2R and R2L,an anomaly detection model based on Support Vector Machine(SVM) and Bayesian classification was suggested.In order to reduce the redundant records in the training data,the BIRCH(Balanced Iterative Reducing and Clustering using Hierarchies) clustering algorithm was used.Besides,the detection model applied SVM for detecting DoS and Probe and used Bayesian classification to detect U2R and R2L.The experimental results show that the proposed model improves obviously the detection rate for U2R and R2L,up to 68.6 percent and 45.7 percent respectively.

An efficient clustering method for fast rendering of time-varying volumetric medical data

Visualization and exploration of time-varying volumetric medical data help clinicians for better diagnosis and treatment. However, it is a challenge to render these data in an interactive manner because of their complexity and large size. We propose an efficient clustering method for fast compression and rendering of these large dynamic data. We divide the volumes into a set of blocks and use a BIRCH (Balanced Iterative Reducing and Clustering using Hierarchies) based algorithm to cluster them, which can usually find a high quality clustering with a single scan of the blocks. In addition, the granularity of clusters can be adaptively adjusted by dynamically configuring threshold values. In each cluster of blocks, a KeyBlock is generated to represent the cluster, and therefore the storage space of the volumes is reduced greatly. In addition, we assign a lifetime to every KeyBlock and implement a dynamic memory management scheme to further reduce the storage space. During the rendering, each KeyBlock is rendered as a KeyImage, which can be reused if the view transformation and transfer function are not changed. This reuse can help to increase the rendering speed significantly. Experimental results showed that the proposed method can achieve good performance in terms of both speed optimization and space reduction.

BIRCH: A New Data Clustering Algorithm and Its Applications

Data clustering is an important technique for exploratory data analysis, and has been studied for several years. It has been shown to be useful in many practical domains such as data classification and image processing. Recently, there has been a growing emphasis on exploratory analysis of very large datasets to discover useful patterns and/or correlations among attributes. This is called data mining, and data clustering is regarded as a particular branch. However existing data clustering methods do not adequately address the problem of processing large datasets with a limited amount of resources (e.g., memory and cpu cycles). So as the dataset size increases, they do not scale up well in terms of memory requirement, running time, and result quality. In this paper, an efficient and scalable data clustering method is proposed, based on a new in-memory data structure called CF-tree, which serves as an in-memory summary of the data distribution. We have implemented it in a system called BIRCH (Balanced Iterative Reducing and Clustering using Hierarchies), and studied its performance extensively in terms of memory requirements, running time, clustering quality, stability and scalability; we also compare it with other available methods. Finally, BIRCH is applied to solve two real-life problems: one is building an iterative and interactive pixel classification tool, and the other is generating the initial codebook for image compression.