Big Datasets Research Articles

An approach to urban system spatial planning in Chengdu Chongqing economic circle using geospatial big data

Evidence suggests city grouping is an important way to implement urbanization in China. However, the Chengdu-Chongqing Economic Circle (CCEC) is a typical dual-core structure, and the development level of each city is different. If we do not focus on the key directions for urban development, it will not be conducive to the new-type urbanization process. Therefore, we use spatial analysis techniques and geographic big data sets to construct an approach for urban system layout optimization from a global perspective. It mainly includes urban extended trend analysis based on night light, multi-modal traffic network analysis, and spatial economic density analysis using Open Street Map (OSM) and Point of Interest (POI) data. The research results show the following interesting findings. Firstly, the historical relationship of cities has a significant impact on city grouping, and efficient transportation connections and prosperous enterprise distribution are key conditions for urban grouping during the acceleration period of urbanization. Secondly, the development of urban grouping should break through administrative restrictions and achieve a moderate separation of administrative divisions and economic divisions. It is beneficial to the rapid growth of the city group and the improvement of the internal structure. Thirdly, the urban group of Southern Sichuan and Western Chongqing (SSWC) is the region with the most potential for growth in CCEC. The urban expansion index (UEI) of the Yibin-Luzhou area is 2.16, and the spatial economic density has increased by 130/km2 in the past decade. Providing flexible development authority and focus on the construction of the Yibin-Luzhou Urban Belt along the Yangtze River is an important way to integrate southern Sichuan. Moreover, we found the UEI has a good universality and it can be used for studying urban expansion trends and city relationships in rapidly developing regions, especially in metropolitan areas or urban agglomeration.

The environmental benefits of dockless bike sharing systems for commuting trips

Using Shanghai, China as an example, we develop a new approach for assessing the environmental benefits of dockless bike sharing system (DLBS), which combines both big data and small data and consider both substitutive modal shifts and complementary modal shifts related to DLBS. With a GPS-based big dataset on DLBS trips and a survey of DLBS commuters, we examine the impacts of the built environment on DLBS usage and modal shift behavior of DLBS commuters. We estimate the environmental benefits of DLBS as the difference in carbon emissions between DLBS trips and their counterfactuals assuming DLBS was unavailable. We find that a DLBS commuting trip on average reduces carbon emission by 185.48 g, much higher than the previous estimates based solely on substitutive modal shifts. The research findings could support policy formation to achieve carbon neutrality by promoting DLBS.

The Essential Facilities Doctrine, Intellectual Property Rights, and Access to Big Data

This paper analyzes the criteria for applying the essential facilities doctrine to intellectual property rights and the possibility of applying it in cases where Big Data is the alleged essential facility. It aims to answer the research question: “What are the specifics of the intellectual property criteria in essential facilities cases and are these criteria applicable to Big Data?” It points to the semantic openness of the “new product” and “technical progress” conditions that have been developed for assessing whether an intellectual property right constitutes an essential facility. The paper argues that the intellectual property criteria are not applicable in all access to Big Data cases because Big Data is not necessarily protected by copyright. While a set of Big Data could be protected by copyright if certain conditions are met, even in such cases the lack of intrinsic value of Big Data significantly limits the applicability of the intellectual property criteria.

ArchaeoSRP

AbstractFor much of its history, archaeological research has relied on site-specific projects, regional comparisons, and theory building from case studies. However, recent research themes concerning the emergence of complex social-ecological systems and long-term land-use legacies require a new approach to archaeological data. Large-scale syntheses of archaeological data provide an effective way forward to address these new research themes. In more concise terms, “big questions” require “big data” to help answer them. The archaeological information collected by the USDA Forest Service is one such “big dataset” and represents an incalculable investment in time, resources, and expertise. This article explores this concept and presents an R package (ArchaeoSRP) designed to extract archaeological information from USDA Forest Service site record files. We demonstrate the functionality of this R package through a case study examining the archaeological data for the Cle Elum Ranger District, within Central Washington's Okanogan-Wenatchee National Forest. Our results reveal the efficiency of using automated methods to extract, organize, and synthesize district-level archaeological data, which, in turn, reveal patterns of precontact and historic land use that were otherwise not distinguishable.

3.C. Skills building seminar: Guidance on using natural experiments to evaluate population health interventions

Abstract There has been a substantial increase in the conduct of natural experimental evaluations since original guidance was published. There have been advances in relevant methodologies, greater availability of big datasets, and a rise in demand for ‘what works’ evidence amenable to natural experimental evaluation. Therefore, it is important that researchers, public health practitioners, funders, and users of intervention research are aware of recent developments and of the strengths, weaknesses, and applicability of natural experimental evaluations. Updated and extended guidance has been developed for using natural experiments to evaluate population health interventions. Guidance was developed from three international workshops and an online consultation with researchers, journal editors and end users of natural experimental evaluations. The project was funded by The UK Medical Research Council and the National Institute for Health and Care Research, and the project team comprised researchers with expertise in natural experimental evaluations. The project had a funder-assigned oversight group and an advisory group of independent experts, including researchers and practitioners with experience of commissioning natural experimental evaluations. This updated and extended guidance provides an integrated guide to the use of natural experimental methods to evaluate population health interventions, offering a range of tools to support its use and detailed, evidence-based information for researchers, funders, publishers, and users of evidence. The objective of this skills building seminar is to upskill public health researchers and practitioners on the design and use of natural experimental methods. The seminar will present key aspects of the guidance and provide the opportunity to discuss and explore each aspect. The first presentation will highlight important considerations when designing and planning a natural experimental evaluation and introduce recommendations for planning, commissioning, conducting, and using evidence from natural experimental evaluations. The second presentation will outline the value and roles of qualitative methods, explaining how qualitative evidence strengthens evaluations. There will then be a discussion and collaborative learning session, providing an opportunity for researchers and public health practitioners to share experiences and engage with the guidance. The discussions will be prompted by examples of natural experimental evaluations relevant to the presentations. Key messages • Natural experimental evaluations are valuable for assessing the impacts of population health interventions. • This workshop provides an opportunity to engage with and discuss new guidance, highlighting good practice in the design, planning, conduct and use of evidence from natural experimental evaluations.

Flexible nonlinear inference and change-point testing of high-dimensional spectral density matrices

This paper studies a flexible approach to analyze high-dimensional nonlinear time series of unconstrained dimension based on linear statistics calculated from spectral average statistics of bilinear forms and nonlinear transformations of lag-window (i.e. band-regularized) spectral density matrix estimators. That class of statistics includes, among others, smoothed periodograms, nonlinear statistics such as coherency, long-run-variance estimators and contrast statistics related to factorial effects as special cases. Especially, we introduce the class of nonlinear spectral averages of the spectral density matrix. Having in mind big data settings, we study a sampling design which includes a sparse sampling scheme. Gaussian approximations with optimal rate are derived for nonlinear time series of growing dimension for these frequency domain statistics and the underlying lag-window (cross-) spectral estimator under non-stationarity. For change-testing (self-standardized) CUSUM statistics are examined. Further, a specific wild bootstrap procedure is proposed to estimate critical values. Simulation studies and an application to SP500 financial returns are provided in a supplement to this paper.

Topic based document modeling for information filtering

Information Filtering (IF), which has been popularly studied in recent years, is one of the areas that applies document retrieval techniques for dealing with the huge amount of information. In IF systems, modelling user’s interest and filtering relevant documents are major parts of the systems. Various approaches have been proposed for modelling the first component. In this study, we utilized a topic-modelling technique, Latent Dirichlet Topic Modelling, to model user’s interest for IFs. In particular, an extended model of it to represent user’s interest named Latent Dirichlet Topic Modelling with high Frequency Occurrences, shorted as LDA_HF, was proposed with the intention to enhance retrieving performance of IFs. The new model was then compared to the existing methods in modelling user’s interest such as BM25, pLSA, and LDA_IF over the big benchmark datasets, RCV1 and R8. The results of extensive experiments showed that the new proposed model outperformed all the state-of-the-art baseline models in user modelling such as BM25, pLSA and LDA_IF according to 4 major measurement metrics including Top20, B/P, MAP, and F1. Hence, the model LDA_HF promises one of the reliable methods of enhancing performance of IFs.

Factorizing social advanced aspects in modern life on human health

The emergence of technology has brought about a dramatic shift in different aspects of life, especially the aspect of societal advancement. As people adapt to such transformation, it is undoubted that their lifestyles adapt as well. However, whether such changes oppose the well-being of an individual and whether there is a relationship between daily habits and health condition is an interesting topic that this research is going to focus on. Meanwhile, the benefits of implementing data analysis have been proven regarding understanding statistical problems. On such account, this work is going to use the powerful method of data analysis into investigating the relationship between lifestyle and health conditions, with the practice of big data sets deploying R programming language. The procedure from gathering to interpreting data is going to be introduced, and the result is then placed in comparison with related existing studies. By doing so, it has shown that the result has reflected accurately the topics and thus, given evidence for the potential of analyzing the connection between habitual practice and health status.

Explainable machine learning models for Medicare fraud detection

As a means of building explainable machine learning models for Big Data, we apply a novel ensemble supervised feature selection technique. The technique is applied to publicly available insurance claims data from the United States public health insurance program, Medicare. We approach Medicare insurance fraud detection as a supervised machine learning task of anomaly detection through the classification of highly imbalanced Big Data. Our objectives for feature selection are to increase efficiency in model training, and to develop more explainable machine learning models for fraud detection. Using two Big Data datasets derived from two different sources of insurance claims data, we demonstrate how our feature selection technique reduces the dimensionality of the datasets by approximately 87.5% without compromising performance. Moreover, the reduction in dimensionality results in machine learning models that are easier to explain, and less prone to overfitting. Therefore, our primary contribution of the exposition of our novel feature selection technique leads to a further contribution to the application domain of automated Medicare insurance fraud detection. We utilize our feature selection technique to provide an explanation of our fraud detection models in terms of the definitions of the selected features. The ensemble supervised feature selection technique we present is flexible in that any collection of machine learning algorithms that maintain a list of feature importance values may be used. Therefore, researchers may easily employ variations of the technique we present.

Slime Mold Optimization with Relational Graph Convolutional Network for Big Data Classification on Apache Spark Environment

Lately, Big Data (BD) classification has become an active research area in different fields namely finance, healthcare, e-commerce, and so on. Feature Selection (FS) is a crucial task for text classification challenges. Text FS aims to characterize documents using the most relevant feature. This method might reduce the dataset size and maximize the efficiency of the machine learning method. Various researcher workers focus on elaborating effective FS techniques. But most of the presented techniques are assessed for smaller datasets and validated by a single machine. As textual data dimensionality becomes high, conventional FS methodologies should be parallelized and improved to manage textual big datasets. This article develops a Slime Mold Optimization based FS with Optimal Relational Graph Convolutional Network (SMOFS-ORGCN) for BD Classification in Apache Spark Environment. The presented SMOFS-ORGCN model mainly focuses on the classification of BD accurately and rapidly. To handle BD, the SMOFS-ORGCN model uses an Apache Spark environment. In the SMOFS-ORGCN model, the SMOFS technique gets executed for reducing the profanity of dimensionality and to improve classification accuracy. In this article, the RGCN technique is employed for BD classification. In addition, Grey Wolf Optimizer (GWO) technique is utilized as a hyperparameter optimizer of the RGCN technique to enhance the classification achievement. To exhibit the better achievement of the SMOFS-ORGCN technique, a far-reaching experiments were conducted. The comparison results reported enhanced outputs of the SMOFS-ORGCN technique over current models.

Improving Data Processing Speed on Large Datasets in a Hadoop Multi-node Cluster using Enhanced Apriori Algorithm

For large data, data mining methods were used on a Hadoop-based distributed infrastructure, using map reduction paradigm approaches for rapid data processing. Though data mining approaches are established methodologies, the Apriori algorithm provides a specific strategy for increasing data processing performance in big data analytics by applying map reduction. Apriori property is used to increase the efficiency of level-wise creation of frequent itemsets by minimizing the search area. A frequent itemset’s subsets must also be frequent (Apriori property). If an itemset is rarely, then all of its supersets are infrequent as well. We refined the apriori approach by varying the degree of order in locating frequent item sets in large clusters using map reduction programming. Fixed Pass Combined Counting (FPC) and Dynamic Pass Combined Counting (DPC) is a classical algorithm which are used for data processing from the huge datasets but their accuracy is not up to the mark. In this article, updated Apriori algorithms such as multiplied-fixed-pass combined counting (MFPC) and average time-based dynamic combined counting (ATDFC) are used to successfully achieve data processing speed. The proposed approaches are based on traditional Apriori core notions in data mining and will be used in the map-reduce multi-pass phase by ignoring pruning in some passes. The optimized-MFPC and optimized-ATDFC map-reduce framework model algorithms were also presented. The results of the experiments reveal that MFPC and ATDFC are more efficient in terms of execution time than previously outmoded approaches such as Fixed Pass Combined Counting (FPC) and Dynamic Pass Combined Counting (DPC). In a Hadoop multi-node cluster, this paradigm accelerates data processing on big data sets. Previous techniques were stated in terms of reducing execution time by 60–80% through the use of several passes. Because of the omitted trimming operation in data pre-processing, our proposed new approaches will save up to 84–90% of that time.

Intrinsic Dynamics of Brain Functional Connectivity in Relation to Women’s Menstrual Cyclic Phases and its Impact on Their Stress, Life Satisfaction, and Social Distress

Research has reported that the fluctuations of sex hormones like estrogen and progesterone lead to neurological changes in the different brain regions across the menstrual cycle and its phases: menses, follicular, and luteal. Although some psychological changes have also been reported for women across their menstrual cycle, not enough studies have approached this topic from a neuropsychological perspective. For this reason, I am investigating brain connectivity fluctuations across a 32-day menstruation cycle and its impact on women’s life satisfaction (LS), perceived stress (PS), and emotional support (ES). For this study, I am using the data (n=406) that has been collected by the Human Connectome Project. I am using an experimental research methodology looking at correlations (connectivity) between brain regions and associated variables of menstrual cycle phases (menses, follicular, and luteal) across a 32-day menstrual cycle, and what this tells us about the participants’ LS, PS, and ES total scores. In neuroscience, artificial intelligence (AI) techniques are applied to understand complex brain functions, brain and behaviour relationships, and shows potential to diagnose neurological disorders and interventions. AI, and elastic net regression, which is a type of statistical technique for neuroimaging data, will be used to analyse the brain activity relationships that I am investigating. In my presentation, I aim to discuss whether functional connectivity in brain regions is linked to women’s LS, PS, and ES, and the implications of this for menstrual health. This pilot study, by using a big data set of relatively healthy participants to further our understanding of this topic, will help to open doors for future research with similar investigating aims for more impactful menstrual related health conditions like Premenstrual Syndrome and Premenstrual Dysphoric Disorder.

Reference-Guided Large-Scale Face Inpainting With Identity and Texture Control

Face inpainting aims at plausibly predicting missing pixels of face images within a corrupted region. Most existing methods rely on generative models learning a face image distribution from a big dataset, which produces uncontrollable results, especially with large-scale missing regions. To introduce strong control for face inpainting, we propose a novel reference-guided face inpainting method that fills the large-scale missing region with identity and texture control guided by a reference face image. However, generating high-quality results under imposing two control signals is challenging. To tackle such difficulty, we propose a dual control one-stage framework that decouples the reference image into two levels for flexible control: High-level identity information and low-level texture information, where the identity information figures out the shape of the face and the texture information depicts the component-aware texture. To synthesize high-quality results, we design two novel modules referred to as Half-AdaIN and Component-Wise Style Injector (CWSI) to inject the two kinds of control information into the inpainting processing. Our method produces realistic results with identity and texture control faithful to reference images. To the best of our knowledge, it is the first work to concurrently apply identity and component-level controls in face inpainting to promise more precise and controllable results.

An approach in using Artificial Intelligence for traffic light optimization (fuzzy method)

Optimization of traffic light has been a hotspot for years, because of technology advancement for the latest 20 years, high demand in international market for car companies to develop, produce more cars and also geographic conditions, high standardatization of live enhance this problem. In this paper I will use an approach in traffic light optimization by using machine learrning technique, to train a set of data, in order to compute and produce best solutions for optimization of traffic light. There have been many methods used such Webster method, Pedri Net algorithm model, fuzzy model and so on. I will use a different approach to fuzzy method, with intention to provide better output result, decrease the amount of released gases in atmosphere, lower delay and waiting time of cars in a traffic jam. In metropol cities consisting million of people where urban infrastructure is complex, only the development and improvement of these methods can make people’s live more simple. One of most early algorithms to minimize cost in travelling through one or lots of routes is Djikstras algorithm, where simultaneous tests to use different routes, chose the best route, thus minimizing consumption and increasing output efficiency. The paper will be divided into several sections: Introduction, where a set of definitions, general terms of traffic light software simulation are presented, Big Data, representing the dataset of input where in Artificial Intelligence are the building ‘bricks’ into comparing and anlazying the ouptut results of different methods/algorithms (Pedri Net algorithm, fuzzy model, improved RNN Djikstra Algorithm etc.). The upcoming section Methodology, gives a general idea of into analyzing Webster’s algorithm and breaking it down into smaller parts, The Derivation of Fuzzy Method, when analyzing bits of components, methodology used in traffic light system and proposed smart system nowadays, fuzzy method is one of the roots in considering clustering techniques, VANet System Architecture, in this section a proposed system architecture is used, it is one of the most importants sections because it approaches a solution, which is a derived form of Internet of Thing (IoT) components into achieving a Smart City System, Result section gives output where in the upcoming sections can be used as proposals and ideas. Neural Network, CNN (Convolution Neural Network), DL (Deep Learning) and RNN (Recurrent Neural Network) are sections dedicated to relation between artificial intelligense and smart city implementation, where the main idea is training large amounts of datasets to propose smart, efficient and reliable solutions. In conlusion and future work section, the proposed solutions underline the importance of correlation between general methods: Fuzzy, Webster method with big datasets (machine learning techniques) and future work ideas highlight the necessity of virtualization and doubling – quadrupling the layers of CNN, which is proportional to hardware computational cost.

ANALYSIS OF EXISTENCE OF RHEUMATOID ARTHRITIS IN GIVEN SUBJECT WITH ASSISTANCE TO INTERPRET RA IMPACT FACTOR DATA

This study proposes an original methodology for distinguishing Rheumatoid Joint pain (RA) utilizing Convolutional Brain Organizations (CNNs) in profound learning. The strategy uses a dataset of clinical pictures, preprocesses them, and feeds them into a CNN model. The CNN design catches pertinent highlights and utilizes completely associated layers for grouping. Execution assessment incorporates precision, awareness, explicitness, and AUC-ROC. Starter results show promising execution, demonstrating the capability of the methodology for early RA discovery. The proposed technique holds guarantee for helping clinical experts in convenient mediation and customized treatment plans. Future work includes approval on bigger datasets and investigating multi-modular information joining for further developed precision and visor. Keywords: Rheumatoid Arthritis, CNN (Convolution Neural Network), Deep Learning

Finding Playing Styles of Badminton Players Using Firefly Algorithm Based Clustering Algorithms

Cluster analysis can be defined as applying clustering algorithms with the goal of finding hidden patterns or groupings in a dataset. Different clustering methods provide different solutions for the same dataset. Traditional clustering algorithms are popular, but handling big data sets is beyond the ability of such methods. We propose three big data clustering methods, based on the Firefly Algorithm (FA). Three different fitness functions were defined on FA using inter cluster distance, intra cluster distance, silhouette value and Calinski-Harabasz Index. The algorithms find the most appropriate cluster centers for a given data set. The algorithms were tested with four popular synthetic data sets and later applied on two badminton data sets to identify different playing styles of players based on physical characteristics. The results specify that the firefly algorithm could generate better clustering results with high accuracy. The algorithms cluster the players to find the most suitable playing strategy for a given player where expert knowledge is needed in labeling the clusters. Comparisons with a PSO based clustering algorithm (APSO) and traditional algorithms point out that the proposed firefly variants work similarly as the APSO method and surpass the performance of traditional algorithms.

When Should They Buy? Surviving Machine Learning Models for Purchase Timing

Unlocking the value of raw data requires transforming it into knowledge and information that drives decision-making. Algorithms for Machine Learning (ML) are capable of analyzing big datasets and delivering accurate predictions. Using machine learning, market segmentation, customer lifetime value, and marketing approaches have been used. This article explores machine learning approaches for marketing, including Support Vector Machines, Genetic Algorithms, Deep Learning, and K-Means. It is rare for ML to be used to forecast when a person will purchase a product or a basket of goods. The survival models Kernel SVM, DeepSurv, Survival Random Forest, and MTLR are evaluated to predict individual tine-purchase choices in this research. Gender, Revenue, Location, Purchase History, Online Behaviour, Interests, and Promotions According to the research, both discounts and CustomerExperience have an impact on the amount of time spent making a purchase. The research indicates that the DeepSurv model best predicted purchase completion. These insights aid in raising conversion rates for marketers.

Group-level brain decoding with deep learning.

Decoding brain imaging data are gaining popularity, with applications in brain-computer interfaces and the study of neural representations. Decoding is typically subject-specific and does not generalise well over subjects, due to high amounts of between subject variability. Techniques that overcome this will not only provide richer neuroscientific insights but also make it possible for group-level models to outperform subject-specific models. Here, we propose a method that uses subject embedding, analogous to word embedding in natural language processing, to learn and exploit the structure in between-subject variability as part of a decoding model, our adaptation of the WaveNet architecture for classification. We apply this to magnetoencephalography data, where 15 subjects viewed 118 different images, with 30 examples per image; to classify images using the entire 1 s window following image presentation. We show that the combination of deep learning and subject embedding is crucial to closing the performance gap between subject- and group-level decoding models. Importantly, group models outperform subject models on low-accuracy subjects (although slightly impair high-accuracy subjects) and can be helpful for initialising subject models. While we have not generally found group-level models to perform better than subject-level models, the performance of group modelling is expected to be even higher with bigger datasets. In order to provide physiological interpretation at the group level, we make use of permutation feature importance. This provides insights into the spatiotemporal and spectral information encoded in the models. All code is available on GitHub (https://github.com/ricsinaruto/MEG-group-decode).

Probabilistic generative transformer language models for generative design of molecules

Self-supervised neural language models have recently found wide applications in the generative design of organic molecules and protein sequences as well as representation learning for downstream structure classification and functional prediction. However, most of the existing deep learning models for molecule design usually require a big dataset and have a black-box architecture, which makes it difficult to interpret their design logic. Here we propose the Generative Molecular Transformer (GMTransformer), a probabilistic neural network model for generative design of molecules. Our model is built on the blank filling language model originally developed for text processing, which has demonstrated unique advantages in learning the “molecules grammars” with high-quality generation, interpretability, and data efficiency. Benchmarked on the MOSES datasets, our models achieve high novelty and Scaf compared to other baselines. The probabilistic generation steps have the potential in tinkering with molecule design due to their capability of recommending how to modify existing molecules with explanation, guided by the learned implicit molecule chemistry. The source code and datasets can be accessed freely at https://github.com/usccolumbia/GMTransformer

Use of Elasticsearch-based business intelligence tools for integration and visualization of biological data.

The emergence of massive datasets exploring the multiple levels of molecular biology has made their analysis and knowledge transfer more complex. Flexible tools to manage big biological datasets could be of great help for standardizing the usage of developed data visualizations and integration methods. Business intelligence (BI) tools have been used in many fields as exploratory tools. They have numerous connectors to link numerous data repositories with a unified graphic interface, offering an overview of data and facilitating interpretation for decision makers. BI tools could be a flexible and user-friendly way of handling molecular biological data with interactive visualizations. However, it is rather uncommon to see such tools used for the exploration of massive and complex datasets in biological fields. We believe that two main obstacles could be the reason. Firstly, we posit that the way to import data into BI tools are not compatible with biological databases. Secondly, BI tools may not be adapted to certain particularities of complex biological data, namely, the size, the variability of datasets and the availability of specialized visualizations. This paper highlights the use of five BI tools (Elastic Kibana, Siren Investigate, Microsoft Power BI, Salesforce Tableau and Apache Superset) onto which the massive data management repository engine called Elasticsearch is compatible. Four case studies will be discussed in which these BI tools were applied on biological datasets with different characteristics. We conclude that the performance of the tools depends on the complexity of the biological questions and the size of the datasets.