Kinds Of Databases Research Articles

Assessment of Compressed and Decompressed ECG Databases for Telecardiology Applying a Convolution Neural Network

Incalculable numbers of patients in hospitals as a result of COVID-19 made the screening of heart patients arduous. Patients who need regular heart monitoring were affected the most. Telecardiology is used for regular remote heart monitoring of such patients. However, the resultant huge electrocardiogram (ECG) data obtained through regular monitoring affects available storage space and transmission bandwidth. These signals can take less space if stored or sent in a compressed form. To recover them at the receiver end, they are decompressed. We have combined telecardiology with automatic ECG arrhythmia classification using CNN and proposed an algorithm named TELecardiology using a Deep Convolution Neural Network (TELDCNN). Discrete cosine transform (DCT), 16-bit quantization, and run length encoding (RLE) were used for compression, and a convolution neural network (CNN) was applied for classification. The database was formed by combining real-time signals (taken from a designed ECG device) with an online database from Physionet. Four kinds of databases were considered and classified. The attained compression ratio was 2.56, and the classification accuracies for compressed and decompressed databases were 0.966 and 0.990, respectively. Comparing the classification performance of compressed and decompressed databases shows that the decompressed signals can classify the arrhythmias more appropriately than their compressed-only form, although at the cost of increased computational time.

Far wall plaque segmentation and area measurement in common and internal carotid artery ultrasound using U-series architectures: An unseen Artificial Intelligence paradigm for stroke risk assessment

Stroke risk assessment using deep learning (DL) requires automated, accurate, and real-time risk assessment while ensuring compact model size. Previous DL paradigms suffered from challenges like memory size, low speed, and complex in nature lacking multi-ethnic, and multi-institution databases. This research segments and measures the area of the plaque far wall of the common carotid (CCA) and internal carotid arteries (ICA) in B-mode ultrasound using four types of solo, namely, UNet, UNet+, UNet++, and UNet+++, and three types of hybrids, namely, Inception-UNet, Fractal-UNet, and Squeeze-UNet, architectures. These seven models are benchmarked against autoencoder-based solution. Three kinds of databases, namely, CCA, ICA, and combined CCA + ICA were implemented using K5 cross-validation protocol. This was validated using unseen Hong Kong data. The CCA database consisted of 379 Japanese images from low-to medium-risk, while the ICA database consisted of 970 Japanese images taken from 97 medium-to high-risk patients. Using the coefficient of correlation (CC) metric between automated measured area and manually delineated area, seven deep learning solo and hybrid models for CCA yielded 0.96, 0.96, 0.98, 0.95, 0.96, and 0.96 respectively, whereas ICA yielded 0.99, 0.99, 0.98, 0.99, 0.98, 0.98, and 0.98 respectively. Area under the receiver operating characteristics curve values for CCA images was 0.97, 0.969, 0.974, 0.969, 0.962, 0.969, and 0.960 respectively, whereas for ICA images were 0.99, 0.989, 0.988, 0.989, 0.986, 0.989, and 0.988, respectively (p < 0.001). The percentage improvement in offline memory size, training time and training parameters for Squeeze-UNet compared to UNet++ were 569%, 122.46%, and 569%, respectively.

Attention-Mechanism-Based Real-Time Gaze Tracking in Natural Scenes With Residual Blocks

Gaze tracking is widely used in fatigue driving detection, eye disease diagnosis, mental illness diagnosis, website or advertising design, virtual reality, gaze-control devices, and human–computer interaction. However, the influence of light, specular reflection, and occlusion, and the change of head pose, especially the ever-changing human pose in natural scenes, have brought great challenges to the accurate gaze tracking. In this article, gaze tracking in natural scenes is studied, and a method based on the convolutional neural network with residual blocks is proposed, in which the attention mechanism is integrated into the network to improve the accuracy of gaze tracking. Furthermore, it is tested on the GazeFollow database, which contains six kinds of databases. The results show that the performance of the proposed method outperforms that of the other state-of-the-art methods in natural scenes. Moreover, the proposed method has better real-time performance and is more suitable for practical applications.

RAPID APPLICATION DEVELOPMENT DALAM RANCANGAN SISTEM INFORMASI PEMINJAMAN DAN PENGEMBALIAN BUKU

Ananda School, which is located in Bogor City, is a school that is very active in the habit of reading its students. This school has a fairly complete library with quality books, but unfortunately, the system of borrowing and returning books is still conventional by recording transactions on many books. This causes problems in the process of searching for books and transactions that have occurred so that they are less effective and efficient in daily activities. The design of the Book Borrowing and Returning Information System is a solution to overcome the problems that occur by using the Rapid Application Development model in the software development method. The data collection process uses the interview method, observations by paying attention to the system that is currently running, and several examples of sites related to information systems for borrowing and returning books, and literacy are also used as reference materials. The Information System for Borrowing and Returning Books at this school has also made it much better with the various kinds of databases used and the reports that are the output of this information system.

DID Code: A Bridge Connecting the Materials Genome Engineering Database with Inheritable Integrated Intelligent Manufacturing

Abstract A data identifier (DID) is an essential tag or label in all kinds of databases—particularly those related to integrated computational materials engineering (ICME), inheritable integrated intelligent manufacturing (I3M), and the Industrial Internet of Things. With the guidance and quick acceleration of the development of advanced materials, as envisioned by official documents worldwide, more investigations are required to construct relative numerical standards for material informatics. This work proposes a universal DID format consisting of a set of build chains, which aligns with the classical form of identifier in both international and national standards, such as ISO/IEC 29168-1:2000, GB/T 27766–2011, GA/T 543.2–2011, GM/T 0006–2012, GJB 7365–2011, SL 325–2014, SL 607–2018, WS 363.2–2011, and QX/T 39–2005. Each build chain is made up of capital letters and numbers, with no symbols. Moreover, the total length of each build chain is not restricted, which follows the formation of the Universal Coded Character Set in the international standard of ISO/IEC 10646. Based on these rules, the proposed DID is flexible and convenient for extending and sharing in and between various cloud-based platforms. Accordingly, classical two-dimensional (2D) codes, including the Hanxin Code, Lots Perception Matrix (LP) Code, Quick Response (QR) code, Grid Matrix (GM) code, and Data Matrix (DM) Code, can be constructed and precisely recognized and/or decoded by either smart phones or specific machines. By utilizing these 2D codes as the fingerprints of a set of data linked with its cloud-based platforms, progress and updates in the composition–processing–structure–property–performance workflow process can be tracked spontaneously, paving a path to accelerate the discovery and manufacture of advanced materials and enhance research productivity, performance, and collaboration.

Development of the Asian Transplant Registry Under the Asian Society of Transplantation

IntroductionIn the Asian region, no international organ transplantation registry exists. Individual centers maintain their own database, or some countries developed a national registration system. To promote collaboration among Asian transplantation societies, the Asian Society of Transplantation (AST) has developed an international transplantation registry for the Asian countries that has been named as the Asian Society Transplant Registry (ASTREG). MethodsIn 2017, the AST council formed a registry committee to develop 2 kinds of databases: ASTREG-N (nationwide level), which collects yearly aggregated data of participating countries, and ASTREG-H (hospital level), which collects the data of transplant recipients and donors from individual centers. ResultsASTREG-N collects each country’s aggregate data of solid-organ transplantation, such as the total number of transplantations and deceased donors. ASTREG-H collects 5 transplant domains, namely recipient baseline characteristics, immunosuppression, post-transplant event, annual post-transplant evaluation, and donor traits. For the ASTREG-H project, South Korea, Philippines, Mongolia, and Myanmar are the current participants. A web-based secure data entry platform with real-time data visualization and automated data verification systems is currently available. Any participating centers can run this platform as their own data collection system. ConclusionThe ASTREG is a collaborative project that will be the representative solid-organ transplantation database in the Asian region. It can aid in the harmonization of transplantation data in the Asian region.

Local distances preserving based manifold learning

In this paper a local manifold learning method based on the local distances preserving (LDP) is proposed. LDP focuses on extracting and preserving the local distances between the data points. In LDP, the coefficients between each data point and its neighbors are calculated based on the inverse of Euclidean distances between them. Then, a minimization function, which is in accordance with the calculated coefficients, is proposed for calculating the embedded data manifold in the low-dimensional representation space. Matching the way of calculation the coefficients and the proposed minimization function is the main significance of the proposed LDP method. In addition, the proposed LDP method shows less sensitivity to the initial parameters such as the number of neighbors and the value of variance. Also, a stochastic based extension of LDP (SLDP) manifold learning method is proposed. The proposed method is compared with the common manifold learning methods based on the achievable recognition rate and the power of the local distances preserving. The experiments have been done on two different kinds of databases: HODA Persian handwritten character database and ORL face image database. The results demonstrate the suitable performance of LDP and SLDP. Also, the results show the robustness of the proposed method to the number of neighbors and the value of variance parameter. Moreover, the proposed method of calculating the embedded data points in LDP and SLDP has less complexity than the similar local manifold learning methods, Laplacian eigenmaps (LEM) and stochastic LEM (SLEM).

Incorporating prior knowledge with simulation data to estimate PSF multipliers using Bayesian logistic regression

Recently, several kinds of databases have been constructed and analyzed to support human reliability analyses. Based on these, some researchers have attempted to model the quantitative relations between performance shaping factors and human error probability. However, the limitations of the traditional regression technique and simulation data employed have come to light. To tackle these issues regarding the traditional statistical analysis, this study proposes an analysis based on the Bayesian logistic regression method that incorporates empirical data with prior knowledge. This method was applied to four different prior knowledge sets and empirical data collected via the Human Reliability data Extraction (HuREX) framework. The mean and credible interval from the obtained posterior distributions were compared with previous research. From the application, we found that the suggested approach is useful in consolidating various data sources to estimate the multipliers of performance shaping factors on error probabilities, producing results robust to the data characteristics, and providing the quantitative uncertainties of the estimation. It is also confirmed that selecting an appropriate prior knowledge and collecting abundant and correct empirical data are important for producing meaningful insights for PSF impacts.

RaaMLab: A MATLAB toolbox that generates amino acid groups and reduced amino acid modes

Amino acid (AA) classification and its different biophysical and chemical characteristics have been widely applied to analyze and predict the structural, functional, expression and interaction profiles of proteins and peptides. We present RaaMLab, a free and open-source MATLAB toolbox, to facilitate studies on proteins and peptides, to generate AA groups and to extract the structural and physicochemical features of reduced AAs (RedAA). This toolbox offers 4 kinds of databases, including the physicochemical properties of AAs and their groupings, 49 AA classification methods and 5 types of biophysicochemical features of RedAAs. These factors can be easily computed based on user-defined alphabet size and AA properties of AA groupings. RaaMLab is an open source freely available at https://github.com/bioinfo0706/RaaMLab. This website also contains a tutorial, extensive documentation and examples.

Customer behavior analysis using real-time data processing

PurposeThe purpose of this paper is to propose customer behavior analysis based on real-time data processing and association rule for digital signage-based online store (DSOS). The real-time data processing based on big data technology (such as NoSQL MongoDB and Apache Kafka) is utilized to handle the vast amount of customer behavior data.Design/methodology/approachIn order to extract customer behavior patterns, customers’ browsing history and transactional data from digital signage (DS) could be used as the input for decision making. First, the authors developed a DSOS and installed it in different locations, so that customers could have the experience of browsing and buying a product. Second, the real-time data processing system gathered customers’ browsing history and transaction data as it occurred. In addition, the authors utilized the association rule to extract useful information from customer behavior, so it may be used by the managers to efficiently enhance the service quality.FindingsFirst, as the number of customers and DS increases, the proposed system was capable of processing a gigantic amount of input data conveniently. Second, the data set showed that as the number of visit and shopping duration increases, the chance of products being purchased also increased. Third, by combining purchasing and browsing data from customers, the association rules from the frequent transaction pattern were achieved. Thus, the products will have a high possibility to be purchased if they are used as recommendations.Research limitations/implicationsThis research empirically supports the theory of association rule that frequent patterns, correlations or causal relationship found in various kinds of databases. The scope of the present study is limited to DSOS, although the findings can be interpreted and generalized in a global business scenario.Practical implicationsThe proposed system is expected to help management in taking decisions such as improving the layout of the DS and providing better product suggestions to the customer.Social implicationsThe proposed system may be utilized to promote green products to the customer, having a positive impact on sustainability.Originality/valueThe key novelty of the present study lies in system development based on big data technology to handle the enormous amounts of data as well as analyzing the customer behavior in real time in the DSOS. The real-time data processing based on big data technology (such as NoSQL MongoDB and Apache Kafka) is used to handle the vast amount of customer behavior data. In addition, the present study proposed association rule to extract useful information from customer behavior. These results can be used for promotion as well as relevant product recommendations to DSOS customers. Besides in today’s changing retail environment, analyzing the customer behavior in real time in DSOS helps to attract and retain customers more efficiently and effectively, and retailers can get a competitive advantage over their competitors.

Add-On Function Development of Design and Analysis System for Hydraulic Support Column Standardization

Background: Hydraulic powered support is a vital equipment of integrated mechanized mining face, used for the protection of people in coal mine underground. Its performance affects the working condition of the whole working face directly. Objective: To improve the its design efficiency and compensate for the lack of experience in designing column, the patent of the Hydraulic Column Design System is designed to provide an optimized model by the expert system according to the operating condition and help designers to finish the cylinder design rapidly and easily. Methods: The patent of analysis system contains a user interface, three kinds of databases and several message responses. The user interface was developed based on the Visual C++ environment. The databases were written in C++ and APDL language which include three databases that are calculation, selection and strength check databases. Besides, to improve the accuracy of design, finite element analysis has been adopted in the system which is run through interface program and completes the stress analysis automatically. Results: Stability and strength analysis on the 8.2m hydraulic support designed by the system showed that the diameters and lengths of the designed cylinders in accord with the national standard and the requirements of movement. The maximum stress exerts on the joint of outer cylinder and middle cylinder and the maximum stress value is 659Mpa which is smaller than the material strength 1170Mpa. Conclusion: The hydraulic column design system is better suited for hydraulic column design. The designed size of column varies from 180mmto 760mm. Meanwhile, by verification, only one point five hours are needed to finish the same work with the system and it can save 94.5% time compared to the traditional method on column design in the same condition effectively. Keywords: ANSYS, design system, finite element analysis, hydraulic support, human-computer interactive code, visual C++.

Feature based local binary pattern for rotation invariant texture classification

The paper presents a novel texture descriptor based on Local Binary Pattern.We address the problems of rotation invariance and illumination invariance.The proposed texture descriptor FbLBP is low-dimension.The results of FbLBP are compared with the state-of-the-art LBP-like variants. The local binary pattern (LBP) descriptor is widely used in texture analysis because of its computational simplicity and robustness to illumination changes. However, LBP has limitations to fully capture discriminative information since only the sign information of the difference vector in a local region is used. To enhance the performance of LBP, we propose a new descriptor for texture classificationfeature based local binary pattern (FbLBP). In the proposed FbLBP, difference vector is decomposed into sign part and magnitude part, the sign part is described by conventional LBP, while the magnitude part is described by two features of the mean and the variance of the magnitude vector. The way we extract magnitude information in difference vector shows high complementarity to the sign part and less sensitive to illumination changes with a low dimensionality. Furthermore, an adaptive local threshold is used to convert these two features into binary codes. The proposed low dimensional FbLBP is very fast to construct and no parameters are required to tune for different kinds of databases. Experimental results on four representative texture databases of Outex, CUReT, UIUC, and XU_HR show that the proposed FbLBP achieves more than 10% improvement compared with conventional LBP and 1%3% improvement compared with the best classification accuracy among other benchmarked state-of-the-art LBP variants.

RACER: accurate and efficient classification based on rule aggregation approach

Rule-based classification is one of the most important topics in the field of data mining due to its wide applications. This article presents a novel rule-based classifier called RACER (Rule Aggregating ClassifiER) to improve the accuracy of data classification. RACER uses a specific rule representation that enables it to consider each instance in the training data as an initial rule, without spending any cost. In order to retrieve an applicable rule set, RACER tries to combine the initial rules together. If the combined rule has a better fitness value in comparison with the two input rules, RACER combines them together. We have used seventeen different datasets from UCI machine learning database repository to evaluate RACER’s capability in classifying various kinds of databases. Moreover, to assess RACER’s performance, we compared our results with some other well-known classifiers including CN.2, PART, C4.5 and SVM. Our experiments show that RACER is an effective classifier in various domains and has better average classification accuracy and understandability in comparison with other applied classifiers.

Variation ontology: annotator guide

BackgroundSystematic representation of information related to genetic and non-genetic variations is required to allow large scale studies, data mining and data integration, and to make it possible to reveal novel relationships between genotype and phenotype. Although lots of variation data is available it is often difficult to use due to lack of systematics.ResultsA novel ontology, Variation Ontology (VariO http://variationontology.org), was developed for annotation of effects, consequences and mechanisms of variations. In this article instructions are provided on how VariO annotations are made. The major levels for description are the three molecules, namely DNA, RNA and protein. They are further divided to four major sublevels: variation type, function, structure, and property, and further up to eight sublevels. VariO annotation summarizes existing knowledge about a variation and its effects and formalizes it so that computational analyses are efficient. The annotations should be made on as many levels as possible. VariO annotations are made in reference to normal states, which vary for each data item including e.g. reference sequences, wild type properties, and activities.ConclusionsDetailed instructions together with examples are provided to indicate how VariO can be used for annotation of variations and their effects. A dedicated tool has been developed for annotation and will be further developed to cover also evidence for the annotations. VariO is suitable for annotation of data in many types of databases. As several different kinds of databases are in a process of adapting VariO annotations it is important to have guidelines to guarantee consistent annotation.

GEOMETRICAL APPROACH TO THE DATA MODELS DESIGN, BASED ON THE GRANULATION THEORY

A different kinds of databases, based on the relational algebra models, are used at present time. In case of graphical databases the known models are need of very large information resources for the data store and data management. In this paper a geometrical approach to the data granulation presented. As a practical result of granulation very sufficient data compression is obtained. The practical application of the new kind of data models is the image databases for the complicated form objects

Secondary Data Analysis of Epidemiology in Asia

It is widely acknowledged that there is great potential for utilizing databases in health care systems in developed countries.1 Modern health care databases have two characteristics that make them attractive to epidemiologists who work on ‘real-world’ problems: routinely conducted data collection and perpetual electronic data storage. Pharmaco-epidemiology and health services research have long taken a leading role in utilizing these kinds of databases in epidemiology, and the emerging research area called comparative effectiveness research (CER) has recently begun to utilize health care databases to weigh the benefits and harms of different interventions. Health care databases are typically used by researchers as secondary data sources, or existing data generated for a purpose different from the research activity for which it is being used.2 There are pros and cons involved in the use of secondary data. From a beneficial aspect, generation of new data is often expensive and time-consuming; further, collection of new data may raise concerns about privacy and unwanted disclosure of data.2 Using secondary data for epidemiologic research can therefore save time and resources if the data meet the study purpose. However, while there are great advantages to the use of secondary data, bias in the database, which can be introduced through confounding, missing data, and misclassification, is a great threat to the validity of epidemiologic research. To address these issues, propensity score matching, sensitivity analysis, and instrumental variable analysis are used to assess and control for bias in observational CER.3 Before applying these novel epidemiologic methods, it is crucial to check the potential for and magnitude of bias by using real-world data. As each database has its own specific purpose (e.g., patient management or health insurance claims), researchers using the databases for secondary purposes sometimes do not know the general characteristics of the databases. The descriptive information provided by databases is therefore important for conducting, analyzing, and interpreting secondary data. Tanihara analyzed the database of health insurance claims (HIC) of the National Health Insurance Organization of Kumamoto Prefecture in May 2010.4 This large database included a total of 3.8 million diagnoses, comprising all computerized health insurance claims in the prefecture for inpatient or outpatient medical care and DPC/PDPS (diagnosis procedure-combination per-diem payment system). Diagnoses in the database were coded in accordance with the International Statistical Classification of Diseases and Related Health Problems, 10th Revision, but some diagnoses were left uncoded. These uncoded diagnoses are often excluded from analyses, which might introduce bias. Tanihara stated that “the main problem of uncoded diagnoses is that the investigation of large administrative databases using HICs in Japan may have been biased if there was a tendency for one or more specific diagnoses to be uncoded more often than others”. He tackled this issue, finding that the overall proportion of uncoded diagnoses was 9.6% and that proportions differed by HIC type and disease category. Among HIC types, DPC/PDPS included more uncoded diagnoses (14.6%) than outpatient (9.3%) and inpatient (10.9%) records. Among disease categories, congenital malformation, deformations, and chromosomal abnormalities (outpatients; 19.5%); injury, poisoning, and certain other consequences of external causes (inpatients; 19.7%); and neoplasms (DPC/PDPS; 77.9%) showed high percentages of uncoded diagnoses. These findings indicate that Japanese health statistics based on computerized health insurance claims might be biased if we exclude uncoded diagnoses from the analysis. Rapid economic growth in the Asia-Pacific region has changed population lifestyle and the environment in which we live over the past decade. Non-communicable diseases (NCDs), such as cancer, cardiovascular disease, and diabetes, are now emerging threats to public health in this region, and effective prevention measures for NCDs are needed. The analysis of ‘real-world’ data is recognized as the best means for development of an evidence-based health policy. Researchers and public health officers in Japan have drawn attention to the national integrated database of medical receipts and health checkup information, called the ‘National Database’. However, access to this database is still restricted under the current government. Miyagawa et al. analyzed part of this database (212 000 residents) and estimated age- and sex-specific prevalence, rates of treatment, and control of cardiovascular risk factors among adults in Shiga prefecture.5 Age- and sex-specific prevalence is a fundamental epidemiologic measure that describes health status by region, and stable estimates can only be obtained from large databases. We hope that descriptive epidemiology involving large databases will become more common in the future. We also encourage researchers to utilize large health care databases available in Asian countries.

Mining Closed Sequential Patterns in Progressive Databases

Previous studies of Mining Closed Sequential Patterns suggested several heuristics and proposed some computationally effective techniques. Like, Bidirectional Extension with closure checking schemas, Back scan search space pruning, and scan skip optimization used in BIDE (BI-Directional Extension) algorithm. Many researchers were inspired with the efficiency of BIDE, have tried to apply the technique implied by BIDE to various kinds of databases; we toofelt that it can be applied over progressive databases. Without tailoring BIDE, it cannot be applied to dynamic databases. The concept of progressive databases explores the nature of incremental databases by defining the parameters like, Period of Interest (POI), user defined minimum support. An algorithm PISA (Progressive mIning Sequential pAttern mining) was proposed by Huang et al. for finding all sequential patterns over progressive databases. The structure of PISA helps in space utilization by limiting the height of the tree, to the length of POI and this issue is also a motivation for further improvement in this work. In this paper, a tree structure LCT (Label, Customer-id, and Time stamp) is proposed, and an approach formining closed sequential patterns using closure checking schemas across the progressive databases concept. The significance of LCT structure is, confining its height to a maximum of two levels. The algorithmic approach describes that the window size can be increased by one unit of time. The complexity of the proposed algorithmic approach is also analysed. The approach is validated using synthetic data sets available in Internet and shows a better performance in comparison to the existing methods.

A New Framework based on Cognitive Psychology for Knowledge Discovery

The mainstream of knowledge discovery encompasses mining algorithms with high-performance and high-scalability in kinds of databases and application backgrounds. Some research were carried out in another shortcut which focuses on research to high-level theory framework in order to reveal the potential essence, principles and complexity of knowledge discovery and then reacts to the mainstream development. The contributions are as follows: 1) the process and system of KD is regards as complex cognitive process and system. 2) Two coordinators, which are used to simulate cognitive psychology characters so as to make system discover knowledge independently and maintain knowledge base in real-time, are proposed. 3) Double bases cooperating mechanism, which can reveal the equivalence relationship between two categories (structure corresponding theorem) in knowledge discovery process under the conditions that database and knowledge base are specially constructed, is proposed. It paves the way for the implementation of directional searching and directional mining of two coordinators. 4) A new process model based on double bases cooperating mechanism is presented. 5) A new mining algorithm is brought forward consequently, which embodies the driving role to mainstream development.

The electronic register patients with hypertensia in Tomsk Region

Within the limits of the regional program «Prevention and treatment of an arterial hypertension for the period of 2004—2008» the electronic register of the patients with hypertensia inTomskRegion has been created. The electronic register is a two-level system where interaction of two kinds of databases is carried out: the first level is the databases of separate medical organization; the second level is the central integrated database. The basic information for the electronic register are documents confirmed by the Health service Ministry of the Russian Federation, that is the coupon of the out-patient patient and a card of dynamic supervision over the patient with hypertensia. All the data about the patients, included in the register are subdivided into unchangeable and changeable ones. The electronic register is an effective control system providing local leading of health service bodies with qualitative and high-grade information in processes of preparation of decision-making and measure taken for prevention and treatment of hypertensia. The electronic register is an effective monitoring system, providing medical authority of important information for taking decisions establishment measures for prevention and treatment of hypertensia.

Auto-Label Threshold Generation for Multiple Relational Classifications based on SOM Network

and Association rule mining are two basic tasks of Data Mining. Classification rules mining finds rules that partition the data into disjoint sets. This paper is based on MrCAR (Multi-relational Classification Algorithm) and Kohonen's Self-Organizing Maps (SOM) approach. SOM is a class of typical artificial neural networks (ANN) with supervised learning which has been widely used in classification tasks. For small disjunction mining, we collocate with a new auto level threshold generation method in our algorithm to solve the problem of unclassified data of MrCAR. So, we optimize the classification rate of MrCAR with SOM network and improve the efficiency of classification. This approach is highly effective for classification of various kinds of databases and has better average classification accuracy in comparison with MrCAR. Finally the results convincingly demonstrated that our proposed algorithm has high accuracy.