Data Analytics Methods Research Articles

A Comprehensive Analysis of Architectures and Methods of Real-Time Big Data Analytics

A Comprehensive Analysis of Architectures and Methods of Real-Time Big Data Analytics

Data Mining and Machine Learning for Condition-based Maintenance

Complex production systems may count thousands of parts and components, subjected to multiple physical and logical connections and interdependencies. This level of complexity inhibits the traditional and statistically-based approach to reliability engineering, failure prediction and maintenance planning. The existing ICT solutions simplify the on-field collection of large amount of data, but require models and tools able to create knowledge from these data. Key questions on how to predict in advance the performance of the production system and the associated failure events could be finally addressed. This paper introduces a set of data analytics models and methods that can be profitably used for decision making in general, and, specifically, in maintenance engineering. These classification models, specifically decision trees, random forests, and neural networks, are applied to a real-world case study, and the resulting accuracy on predicting faults is quantified and compared. We used the historical profiles of the energy variables of an high-speed packaging machine to find out some strategies for the prediction of a given failure. The conducted experiments demonstrate that the accuracy of the random forest is slightly better than the other methods, but even increases the probability of false alarm, which would result in unwanted production break-down. Even though the obtained results are promising, they leave room for further experiments based on the application of other classifiers, rather than the definition of customized methods able to embrace such complexity.

Social evaluation of innovative drugs: A method based on big data analytics

The evaluation of drugs is a professional and time consuming process which involves a series of clinical trials and evidence-based verifications. However, an innovative drug may still suffer from unpredictable risks after coming into market due to the complex circumstances in practical utilization. Owing to the popularization of information networks and social media, big data analytics exhibits a new perspective of social evaluation as the supplementary means on this issue. This paper designed a Hadoop platform for data collection and processing, and explored the social evaluation of innovative drugs based on big data analytics. Through the analysis of mined data and affective computing on online comments, a new Chinese drug extracted from marine organisms can be evaluated comprehensively by the proposed method. Furthermore, the potential utilization of fullerene materials may be considered for improving its curative effects. Research work of this paper provides a big data analytics method for social evaluation of innovative drugs as well as their promising improvements.

Data Analytics based Deep Mayo Predictor for IPL-9

This paper presents a Deep Mayo Predictor model for predicting the outcomes of the matches in IPL 9 being played in April – May, 2016. The model has three components which are based on multifarious considerations emerging out of a deeper analysis of T20 cricket. The models are created using Data Analytics methods from machine learning domain. The prediction accuracy obtained is high as the Mayo Predictor Model is able to correctly predict the outcomes of 39 matches out of the 56 matches played in the league stage of the IPL IX tournament. Further improvement in the model can be attempted by using a larger training data set than the one that has been utilized in this work. No such effort at creating predictor models for cricket matches has been reported in the literature.

Disease networks and predictive methods for clinical data analytics

Disease networks and predictive methods for clinical data analytics

Big data analytics in bioinformatics: architectures, techniques, tools and issues

Bioinformatics research is characterized by voluminous and incremental datasets and complex data analytics methods. The machine learning methods used in bioinformatics are iterative and parallel. These methods can be scaled to handle big data using the distributed and parallel computing technologies. Usually big data tools perform computation in batch mode and are not optimized for iterative processing and high data dependency among operations. In the recent years, parallel, incremental, and multi-view machine learning algorithms have been proposed. Similarly, graph-based architectures and in-memory big data tools have been developed to minimize I/O cost and optimize iterative processing. However, standard big data architectures are still lacking. Also appropriate tools are not available for many important bioinformatics problems, such as fast construction of co-expression and regulatory networks and salient module identification, detection of complexes over growing protein-protein interaction data, fast analysis of massive DNA, RNA, and protein sequence data, and fast querying on incremental and heterogeneous disease networks. This paper addresses the issues and challenges posed by several big data problems in bioinformatics, and gives an overview of the state of the art and the future research opportunities.

Big Data Analytics Methods and Automotive Specific Modulations

Big Data Analytics Methods and Automotive Specific Modulations

DESIGNING INFORMATION VISUALIZATION FOR HIGHER EDUCATION INSTITUTIONS: A PRE-DESIGN STUDY

The multidimensional of students’ data and the limitations of the currently-used data management tools in higher education institutions (HEIs) have been identified as causes of constrained decision-making process in the domain.This, therefore, necessitates a pre-design study for the HEI students’ datafocused InfoVis.The objectives of this study are to identify the prevailing practices of HEI data management, the data analytics methods that are generally employed by HEI data analysts and the comprehensive dimensions that are related to HEI students and adequate for conveying the domain explicit knowledge preferences. A mixed method approach with survey questionnaire and interview as data collection methods, is used.The contributions of this study, among others, are: (i) identification of the pattern and relationship of the domains explicit knowledge preferences; and (ii) the elicitation and description of students’ data dimensions.These are expected to form the basis for the choice and implementation of the content delivery techniques in designing the domain-focused InfoVis.Our future works therefore entail developing the HEI InfoVis conceptual framework, designing the HEI students’ data-focused InfoVis and conducting its users’ experimental evaluation.

Survey Big Data Analytics, Applications and Privacy Concerns

Background/Objectives: The sources of big data are social media, enterprise data, unstructured data, sensor and clickstream data. The objective is to integrate this variety of data at one platform for processing the big data and find privacy concerns. Methods: The privacy concerns are raised due to unauthorized data extraction, collection and sharing information about user. For integrating and processing of big data; different tools and techniques are available. Findings: General framework for privacy preserving is discussed. Advancements in the big data analytics methods have posed different challenges in front of user. Due to large volume and variety of big data many organizations cannot process the data and needs to outsource it. While sharing such data for processing; there is need to apply proper privacy preserving measures. Application/Improvements: Privacy preserving techniques have applications in electronic health record processing, government surveys, outsourcing enterprise data for processing.

From Big to Little Data for Natural Disaster Recovery: How Online and On-the-Ground Activities are Connected?

Following a major natural disaster, many turn to social media to communicate about their situation and try to seek help in disaster recovery. With millions of social media posts, it can be difficult for disaster management organizations to tap into these immense social networks to find the data needed and to connect individuals to networks that can provide assistance. This study takes big data analytics methods and applies them to a specific context, examining how active and influential members of Facebook groups aided in disaster recovery following Hurricane Sandy. It uses network analysis methods for finding influential members and a web-survey for learning about their background and volunteer activity inside and outside of their Facebook groups. The findings show that the majority of the active online members are also actively involved in on-the-ground recovery activities. They also have the capacity and willingness to work as volunteers. These members have important roles in the integration of online and on-the-ground disaster recovery efforts. Local governments and disaster management organizations should be prepared to incorporate social media data in their formal disaster recovery processes. This incorporation requires the integration of big data analysis methods with social science theories and methods.

Ascertainment-adjusted parameter estimation approach to improve robustness against misspecification of health monitoring methods

Condition monitoring aims at ensuring system safety which is a fundamental requirement for industrial applications and that has become an inescapable social demand. This objective is attained by instrumenting the system and developing data analytics methods such as statistical models able to turn data into relevant knowledge. One difficulty is to be able to correctly estimate the parameters of those methods based on time-series data. This paper suggests the use of the Weighted Distribution Theory together with the Expectation–Maximization algorithm to improve parameter estimation in statistical models with latent variables with an application to health monotonic under uncertainty. The improvement of estimates is made possible by incorporating uncertain and possibly noisy prior knowledge on latent variables in a sound manner. The latent variables are exploited to build a degradation model of dynamical system represented as a sequence of discrete states. Examples on Gaussian Mixture Models, Hidden Markov Models (HMM) with discrete and continuous outputs are presented on both simulated data and benchmarks using the turbofan engine datasets. A focus on the application of a discrete HMM to health monitoring under uncertainty allows to emphasize the interest of the proposed approach in presence of different operating conditions and fault modes. It is shown that the proposed model depicts high robustness in presence of noisy and uncertain prior.

A comparative study of data fusion for RGB-D based visual recognition

Data fusion from different modalities has been extensively studied for a better understanding of multimedia contents. On one hand, the emergence of new devices and decreasing storage costs cause growing amounts of data being collected. Though bigger data makes it easier to mine information, methods for big data analytics are not well investigated. On the other hand, new machine learning techniques, such as deep learning, have been shown to be one of the key elements in achieving state-of-the-art inference performances in a variety of applications. Therefore, some of the old questions in data fusion are in need to be addressed again for these new changes. These questions are: What is the most effective way to combine data for various modalities? Does the fusion method affect the performance with different classifiers? To answer these questions, in this paper, we present a comparative study for evaluating early and late fusion schemes with several types of SVM and deep learning classifiers on two challenging RGB-D based visual recognition tasks: hand gesture recognition and generic object recognition. The findings from this study provide useful policy and practical guidance for the development of visual recognition systems.

Prescriptive Control of Business Processes

This paper proposes a concept for a prescriptive control of business processes by using event-based process predictions. In this regard, it explores new potentials through the application of predictive analytics to big data while focusing on production planning and control in the context of the process manufacturing industry. This type of industry is an adequate application domain for the conceived concept, since it features several characteristics that are opposed to conventional industries such as assembling ones. These specifics include divergent and cyclic material flows, high diversity in end products’ qualities, as well as non-linear production processes that are not fully controllable. Based on a case study of a German steel producing company – a typical example of the process industry – the work at hand outlines which data becomes available when using state-of-the-art sensor technology and thus providing the required basis to realize the proposed concept. However, a consideration of the data size reveals that dedicated methods of big data analytics are required to tap the full potential of this data. Consequently, the paper derives seven requirements that need to be addressed for a successful implementation of the concept. Additionally, the paper proposes a generic architecture of prescriptive enterprise systems. This architecture comprises five building blocks of a system that is capable to detect complex event patterns within a multi-sensor environment, to correlate them with historical data and to calculate predictions that are finally used to recommend the best course of action during process execution in order to minimize or maximize certain key performance indicators.

INTEGRATING DATA ANALYTICS AND SIMULATION METHODS TO SUPPORT MANUFACTURING DECISION MAKING.

Modern manufacturing systems are installed with smart devices such as sensors that monitor system performance and collect data to manage uncertainties in their operations. However, multiple parameters and variables affect system performance, making it impossible for a human to make informed decisions without systematic methodologies and tools. Further, the large volume and variety of streaming data collected is beyond simulation analysis alone. Simulation models are run with well-prepared data. Novel approaches, combining different methods, are needed to use this data for making guided decisions. This paper proposes a methodology whereby parameters that most affect system performance are extracted from the data using data analytics methods. These parameters are used to develop scenarios for simulation inputs; system optimizations are performed on simulation data outputs. A case study of a machine shop demonstrates the proposed methodology. This paper also reviews candidate standards for data collection, simulation, and systems interfaces.

Big-deep-smart data in imaging for guiding materials design.

Harnessing big data, deep data, and smart data from state-of-the-art imaging might accelerate the design and realization of advanced functional materials. Here we discuss new opportunities in materials design enabled by the availability of big data in imaging and data analytics approaches, including their limitations, in material systems of practical interest. We specifically focus on how these tools might help realize new discoveries in a timely manner. Such methodologies are particularly appropriate to explore in light of continued improvements in atomistic imaging, modelling and data analytics methods.

A Review of Emerging Technologies for the Management of Diabetes Mellitus.

High prevalence of diabetes mellitus (DM) along with the poor health outcomes and the escalated costs of treatment and care poses the need to focus on prevention, early detection and improved management of the disease. The aim of this paper is to present and discuss the latest accomplishments in sensors for glucose and lifestyle monitoring along with clinical decision support systems (CDSSs) facilitating self-disease management and supporting healthcare professionals in decision making. A critical literature review analysis is conducted focusing on advances in: 1) sensors for physiological and lifestyle monitoring, 2) models and molecular biomarkers for predicting the onset and assessing the progress of DM, and 3) modeling and control methods for regulating glucose levels. Glucose and lifestyle sensing technologies are continuously evolving with current research focusing on the development of noninvasive sensors for accurate glucose monitoring. A wide range of modeling, classification, clustering, and control approaches have been deployed for the development of the CDSS for diabetes management. Sophisticated multiscale, multilevel modeling frameworks taking into account information from behavioral down to molecular level are necessary to reveal correlations and patterns indicating the onset and evolution of DM. Integration of data originating from sensor-based systems and electronic health records combined with smart data analytics methods and powerful user centered approaches enable the shift toward preventive, predictive, personalized, and participatory diabetes care. The potential of sensing and predictive modeling approaches toward improving diabetes management is highlighted and related challenges are identified.

Secondary Data Analytics of Aquaporin Expression Levels in Glioblastoma Stem-Like Cells.

Glioblastoma is the most common brain tumor in adults in which recurrence has been attributed to the presence of cancer stem cells in a hypoxic microenvironment. On the basis of tumor formation in vivo and growth type in vitro, two published microarray gene expression profiling studies grouped nine glioblastoma stem-like (GS) cell lines into one of two groups: full (GSf) or restricted (GSr) stem-like phenotypes. Aquaporin-1 (AQP1) and aquaporin-4 (AQP4) are water transport proteins that are highly expressed in primary glial-derived tumors. However, the expression levels of AQP1 and AQP4 have not been previously described in a panel of 92 glioma samples. Therefore, we designed secondary data analytics methods to determine the expression levels of AQP1 and AQP4 in GS cell lines and glioblastoma neurospheres. Our investigation also included a total of 2,566 expression levels from 28 Affymetrix microarray probe sets encoding 13 human aquaporins (AQP0–AQP12); CXCR4 (the receptor for stromal cell derived factor-1 [SDF-1], a potential glioma stem cell therapeutic target]); and PROM1 (gene encoding CD133, the widely used glioma stem cell marker). Interactive visual representation designs for integrating phenotypic features and expression levels revealed that inverse expression levels of AQP1 and AQP4 correlate with distinct phenotypes in a set of cell lines grouped into full and restricted stem-like phenotypes. Discriminant function analysis further revealed that AQP1 and AQP4 expression are better predictors for tumor formation and growth types in glioblastoma stem-like cells than are CXCR4 and PROM1. Future investigations are needed to characterize the molecular mechanisms for inverse expression levels of AQP1 and AQP4 in the glioblastoma stem-like neurospheres.

A Survey of Big Data Analytics in Healthcare and Government

Abstract This paper gives an insight of how we can uncover additional value from the data generated by healthcare and government. Large amount of heterogeneous data is generated by these agencies. But without proper data analytics methods these data became useless. Big Data Analytics using Hadoop plays an effective role in performing meaningful real-time analysis on the huge volume of data and able to predict the emergency situations before it happens. It describes about the big data use cases in healthcare and government.

Identifying key variables in African American adherence to colorectal cancer screening: the application of data mining.

BackgroundThis paper reports on an effort to identify a streamlined set of issues important for colorectal cancer communication and interventions with older African Americans.MethodsAfrican American (N = 1,021), 683 women and 338 men, 50 to 75 years completed a telephone survey addressing demographics, colorectal cancer screening, cancer attitudes, and cancer related cultural attitudes. Several data analytics methods were applied and evaluated. Among them, results from associative data mining identified key variables and logistic regression was used to confirm associations to screening adherence.ResultsSets of co-occurring variables identified by associative data mining methods are extracted to further study differences between adherent and non-adherent groups. Logistic regressions suggested four variables were significantly associated with adherence: healthcare provider colonoscopy recommendation, prevention services at the place health care is usually sought, a history of colitis, and a history of polyps.ConclusionsThe findings suggest a streamlined set of issues and concerns that may be used by providers advising patients or developing colorectal cancer intervention strategies for older African Americans. The data suggest the continued importance of healthcare provider recommendation to screen. It is important that providers give a clear recommendation to screen regardless of the test ultimately selected and should advise all patients that family history and the absence of symptoms or colitis do not eliminate the value of screening.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2458-14-1173) contains supplementary material, which is available to authorized users.