Data Setting Research Articles

Can educational attainment promote renewable energy consumption? Evidence from heterogeneous panel models

PurposeFor mitigating climate change, renewable energy consumption is recognized as one of the policy measures worldwide. However, there is a dearth of empirical studies focusing on education as one of the determinants of renewable energy consumption in the existing literature. Thus, this study aims to explore the impact of education, economic growth and foreign direct investment, financial development, CO2 emissions and urbanization on renewable energy consumption.Design/methodology/approachThis study considers a balanced panel of selected South Asian Association of Regional Cooperation (SAARC) countries, namely, India, Pakistan, Sri Lanka, Nepal and Bangladesh, during the period 1995–2015. The study uses sophisticated second-generation panel data models for empirical analysis.FindingsThe result reveals that education and economic growth significantly enhance renewable energy consumption, whereas foreign direct investment, financial development, CO2 emissions and urbanization reduce it. Further, unidirectional causality from education, economic growth and urbanization to renewable energy consumption was observed, whereas a bidirectional causality was found between renewable energy consumption and financial development.Practical implicationsThe emanated finding of this study is supposed to be helpful for the environmentalists, economists, banking sector and the practitioners in urban development can take insights from the study while framing the energy policy.Originality/valueThis is the first study that examines the role of education on renewable energy consumption in heterogeneous panel data settings for the selected SAARC countries.

From viral evolution to spatial contagion: a biologically modulated Hawkes model.

Mutations sometimes increase contagiousness for evolving pathogens. During an epidemic, scientists use viral genome data to infer a shared evolutionary history and connect this history to geographic spread. We propose a model that directly relates a pathogen's evolution to its spatial contagion dynamics-effectively combining the two epidemiological paradigms of phylogenetic inference and self-exciting process modeling-and apply this phylogenetic Hawkes process to a Bayesian analysis of 23 421 viral cases from the 2014 to 2016 Ebola outbreak in West Africa. The proposed model is able to detect individual viruses with significantly elevated rates of spatiotemporal propagation for a subset of 1610 samples that provide genome data. Finally, to facilitate model application in big data settings, we develop massively parallel implementations for the gradient and Hessian of the log-likelihood and apply our high-performance computing framework within an adaptively pre-conditioned Hamiltonian Monte Carlo routine. Supplementary data are available at Bioinformatics online.

Central subspaces review: methods and applications

Central subspaces have long been a key concept for sufficient dimension reduction. Initially constructed for solving problems in the p<n setting, central subspace methods have seen many successes and developments. However, over the last few years and with the advancement of technology, many statistical problems are now situated in the high dimensional setting where p>n. In this article we review the theory of central subspaces and give an updated overview of central subspace methods for the p≤n, p>n and big data settings. We also develop a new classification system for these techniques and list some R and MATLAB packages that can be used for estimating the central subspace. Finally, we develop a central subspace framework for bioinformatics applications and show, using two distinct data sets, how this framework can be applied in practice.

Mapping Temporary Slums From Satellite Imagery Using a Semi-Supervised Approach

One billion people worldwide are estimated to be living in slums, and documenting and analyzing these regions is a challenging task. As compared to regular slums; the small, scattered and temporary nature of temporary slums makes data collection and labeling tedious and time-consuming. To tackle this challenging problem of temporary slums detection, we present a semi-supervised deep learning segmentation-based approach; with the strategy to detect initial seed images in the zero-labeled data settings. A small set of seed samples (32 in our case) are automatically discovered by analyzing the temporal changes, which are manually labeled to train a segmentation and representation learning module. The segmentation module gathers high dimensional image representations, and the representation learning module transforms image representations into embedding vectors. After that, a scoring module uses the embedding vectors to sample images from a large pool of unlabeled images and generates pseudo-labels for the sampled images. These sampled images with their pseudo-labels are added to the training set to update the segmentation and representation learning modules iteratively. To analyze the effectiveness of our technique, we construct a large geographically marked dataset of temporary slums. This dataset constitutes more than 200 potential temporary slum locations (2.28 square kilometers) found by sieving sixty-eight thousand images from 12 metropolitan cities of Pakistan covering 8000 square kilometers. Furthermore, our proposed method outperforms several competitive semi-supervised semantic segmentation baselines on a similar setting. The code and the dataset will be made publicly available.

Supervised Change Detection Using Prechange Optical-SAR and Postchange SAR Data

Change detection using satellite/aerial images is used to quantify the impacts of many natural and man-made disasters. At the occurrence of such events, both pre-change optical and Synthetic Aperture Radar (SAR) images can be obtained by going back in time. However, the availability of the post-change optical image is often hindered by the presence of artifacts like clouds. To circumnavigate this, we propose a novel change detection data setting which uses both optical and SAR images pre-change, yet only SAR imagery post-change. For this challenging scenario, we propose a Siamese network that processes the pre-change and post-change SAR inputs using a shared set of weights, while the pre-change optical input is processed using a network that do not share the weights with the SAR inputs. The encoded weights from the three networks are fused and finally decoded using a common decoder to obtain the change map. Our model effectively fuses multi-sensor information and can obtain satisfactory result despite the absence of the post-change optical image. Experimental results on a multi-sensor urban dataset demonstrate the effectiveness of the proposed approach. Code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://gitlab.lrz.de/ai4eo/cd/-/tree/main/optSarSarCd</uri>

Developing data capability with non-profit organisations using participatory methods

In this paper, we explore the methodologies underpinning two participatory research collaborations with Australian non-profit organisations that aimed to build data capability and social benefit in data use. We suggest that studying and intervening in data practices in situ, that is, in organisational data settings expands opportunities for improving the social value of data. These situated and collaborative approaches not only address the ‘expertise lag’ for non-profits but also help to realign the potential social value of organisational data use. We explore the relationship between data literacy, data expertise and data capability to test the idea that collaborative work with non-profit organisations can be a practical step towards addressing data equity and generating data-driven social outcomes. Rather than adopting approaches to data literacy that focus on individuals – or ideal ‘data citizens’ – we target the organisation-wide data settings, goals and practices of the non-profit sector. We conclude that participatory methods can embed social value-generating data capability where it can be sustained at an organisational level, aligning with community needs to promote collaborative data action.

可持续景观规划——融合景观可持续性研究与地理设计

人类活动对景观的影响遍及世界各个角落, 其广度、强度、频度不断增大。为了改变生态环境在人类社会经济增长压力下的不可持续状态, 景观需要科学合理的设计与管理。可持续景观规划是在景观尺度上, 在可持续性科学的指导下, 平衡生态系统服务的供需关系, 将生态、社会、经济活动过程反映到空间优化上, 不断提高人类福祉的规划过程。为了能够实践该规划思路, 在景观可持续性研究与地理设计融合的概念框架基础上, 提出可持续景观规划的八个步骤, 在问题诊断与目标设定、多源数据组织与管理、尺度匹配与多尺度分析、景观格局过程分析和可持续性评估、生态系统服务与景观格局关系模拟、情景分析与方案评估、地理设计平台搭建、可视化与人机互动等关键步骤中, 融入了强-弱可持续性、多尺度分析、生态系统服务、可持续性指标、大数据应用, 以及文化和地方感。研究提出的可持续性景观规划步骤与实现路径体现了多学科交叉、实时评估反馈、信息技术应用、利益相关者参与的特点, 可被用于旨在提高人类福祉的多尺度空间优化和可持续景观设计。;The scope, intensity, and frequency of human influences on landscapes have been increasing rapidly all over the world. In order to change the unsustainable state of the environment which is pressured by human activities, landscapes need to be planned and managed properly following scientific principles. Sustainable landscape planning is a sustainability science-based planning process that aims to balance the supply and demand of ecosystem services at the landscape scale, spatially optimizes ecological, social and economic processes, and ultimately improves human wellbeing. In order to implement this planning idea, here we propose an eight-step framework of sustainable landscape planning, which integrates landscape sustainability research and GeoDesign. Strong/weak sustainability, multiple scales, ecosystem services, sustainability indicators, big data application, and culture and sense of place are integrated into the key steps of problem diagnosis and goal setting, organization and management of multisource data, scale matching and multiscale analysis, landscape pattern-process analysis and sustainability assessment, landscape pattern-ecosystem services relationship simulation, scenario analysis and scheme evaluation, GeoDesign platform construction, and visualization and human-computer interaction. The procedural steps and implementation pathways of sustainable landscape planning proposed in this study are characterized by interdisciplinarity, real-time evaluation feedbacks, use of advanced information technologies, and stakeholder participation, which can be used for multi-scale spatial optimization and sustainable landscape design to enhance human wellbeing.

Efficient and Secure Multi-Keyword Ranked Search and Group Data Sharing for Encrypted Cloud Data

Nowadays due to the popularity of clouds and their quality services, many of the users such as organizations, industries as well as individuals are migrating towards the cloud to store their important, confidential data and get easy access to their data anywhere anytime over the internet. Different encryption techniques are employed in cloud computing to ensure data confidentiality, security, and privacy. This makes it harder for endusers to retrieve precise data. Due to the huge amount of data over the cloud and multiple data users, secure data storage and retrieval are required. So in this paper develop an efficient data group sharing and multi-keyword ranked search method for encrypted cloud data collection in this research work. The developed system is implemented using the El-Gamal cryptography algorithm to provide security through effective key generation techniques and encryption strategy. Here, a multi-owner data setting is used instead of a centralized data owner setting; each member of the system in one particular group gets equal rights for both searching and sharing functionality and this may increase system usability. By taking into consideration lots of data in the cloud, the vector space model and TF-IDF model are utilized and according to the cosine similarity score, the method generates a ranked multikeyword search result to deliver effective query result from numerous data and enhance secrecy in the situation of numerous data owners. In this system searching efficiency is improved by developing an index-based search structure. In the group, data can disperse with co-owners/users by developing a role-based access policy (RBAP), and a user revocation strategy is developed with low computation time as well as communication overhead. At last, the efficiency and security of a developed system are exhibited by broad exploratory assessment.

Linearized inverse Schrödinger potential problem with partial data and its deep neural network inversion

&lt;p style='text-indent:20px;'&gt;We study the linearized inverse Schrödinger potential problem with (many) partial boundary data. By fixing specific partial boundary these measurements are realized by the linearized local Dirichlet-to-Neumann map. When the wavenumber is assumed to be large, we verify a Hölder type increasing stability by constructing the complex exponential solutions in a reflection form. Meanwhile, the linearized inverse Schrödinger potential problem admits an integral equation where the unknown potential function is indirectly contained there. Such a formulation allows us to adopt a deep neural network inversion algorithm. Numerical examples show that one can reconstruct the unknown potential function stably within the partial boundary data setting.&lt;/p&gt;

Privacy-preserving harmonization via distributed ComBat

Challenges in clinical data sharing and the need to protect data privacy have led to the development and popularization of methods that do not require directly transferring patient data. In neuroimaging, integration of data across multiple institutions also introduces unwanted biases driven by scanner differences. These scanner effects have been shown by several research groups to severely affect downstream analyses. To facilitate the need of removing scanner effects in a distributed data setting, we introduce distributed ComBat, an adaptation of a popular harmonization method for multivariate data that borrows information across features. We present our fast and simple distributed algorithm and show that it yields equivalent results using data from the Alzheimer’s Disease Neuroimaging Initiative. Our method enables harmonization while ensuring maximal privacy protection, thus facilitating a broad range of downstream analyses in functional and structural imaging studies.

Trends of blood pressure control in the U.S. during the COVID-19 pandemic

ImportanceCOVID-19 altered lifestyles and disrupted routine health care. Whether blood pressure (BP) control worsened during COVID-19 is unknown. ObjectiveTo understand whether home BP control worsened during COVID-19 across the United States (US) . Design, Setting, and ParticipantsA population-based analysis of home BP data from 72,706 participants enrolled in a digital health hypertension control program. Data was compared before (January 2019 to March 2020) and during (April 2020 to August 2020) COVID-19. Main Outcomes and MeasuresMonthly mean home BP readings, systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP) were quantified before and during the pandemic. Multivariable adjustments were made for age, sex, race, region, and months enrolled. Home BP readings were also classified based on monthly averages and highest home BP readings into risk groups: Stage 2 HTN: BP> = 135 or DBP> = 85; Uncontrolled HTN: SBP> = 145 or DBP> = 95; or Severely uncontrolled HTN: SBP> = 160 or DBP> = 100). ResultsOverall, 72,706 participants were enrolled in a digital health hypertension program between 1/1/2019 and 8/31/2020. Compared with participants pre-COVID-19 (n = 33,440), those during COVID-19 (n = 39,266) were of similar age (mean 53.0 ± 10.7 years vs 53.3 ± 10.8 years); sex (46% vs 50.6% female) and race (29.1% vs 34.2% non-white). Relative to pre-Covid (Apr-Aug 2019) the mean monthly number of home BP readings rose during COVID-19 (Apr-Aug, 2020), from 7.3 to 9.3 per month (P < .001). During COVID-19, participants had higher monthly adjusted mean SBP (131.6 mmHg vs. 127.5 mmHg, P < .001); DBP (80.2 mmHg vs. 79.2 mmHg, P < .001); and MAP (97.4 mmHg vs. 95.3 mmHg; P < .001). Relative to the pre-pandemic period, during COVID-19 the proportion of participants with a mean monthly BP classified as uncontrolled or severely uncontrolled hypertension also rose, 15% vs 19% and 4% vs 5%, respectively Conclusions and RelevanceBased on home BP readings, mean monthly BP rose modestly after COVID-19, despite increased utilization of home monitoring. Further studies are needed to examine the longitudinal effects of the pandemic on cardiovascular disease risk factors, the impact of these on long-term population health.

Transfer learning for spatio-temporal transferability of real-time crash prediction models

Real-time crash prediction is a heavily studied area given their potential applications in proactive traffic safety management in which a plethora of statistical and machine learning (ML) models have been developed to predict traffic crashes in real-time. However, one of the fundamental issues relating to the application of these models is spatio-temporal transferability. The present paper attempts to address this gap of knowledge by combining Generative Adversarial Network (GAN) and transfer learning to examine the transferability of real-time crash prediction models under an extremely imbalanced data setting. Initially, a baseline model was developed using Deep Neural Network (DNN) with crash and microscopic traffic data collected from M1 Motorway in the UK in 2017. The dataset utilised in the baseline model is naturally imbalanced with 257 crash cases and 16,359,163 non-crash cases. To overcome data imbalance issue, Wasserstein GAN (WGAN) was utilised to generate synthetic crash data. Non-crash data were randomly undersampled due to computational limitations. The calibrated model was then applied to predict traffic crashes for five other datasets obtained from M1 (2018), M4 (2017 & 2018 separately) and M6 Motorway (2017 & 2018 separately) by using transfer learning. Model transferability was compared with standalone models and direct transfer from the baseline model. The study revealed that direct transfer is not feasible. However, models become transferable temporally, spatially, and spatio-temporally if transfer learning is applied. The predictability of the transferred models outperformed existing studies by achieving high Area Under Curve (AUC) values ranging between 0.69 and 0.95. The best transferred model can predict nearly 95% crashes with only a 5% false alarm rate by tuning thresholds. Furthermore, the performances of transferred models are on par with or better than the standalone model. The findings of this study proves that transfer learning can improve model transferability under extremely imbalanced settings which helps traffic engineers in developing highly transferable models in future.

Survival outcomes of management in metastatic gastric adenocarcinoma patients

Chemotherapy is generally considered as the main treatment for metastatic gastric adenocarcinoma. The role of gastrectomy for metastatic gastric cancer without obvious symptoms is controversial. The objective of this study is to investigate survival outcomes of treatment modalities using a real-world data setting. A retrospective cohort study was designed using the Taiwan Cancer Registry database. We identified the treatment modalities and used Kaplan–Meier estimates and Cox regressions to compare patient survival outcomes. From 2008 to 2015, 5599 gastric adenocarcinoma patients were diagnosed with metastatic disease (M1). The median overall survival (OS) of patients with surgery plus chemotherapy had the longest survival of 14.2 months. The median OS of the patients who received chemotherapy alone or surgery alone was 7.0 and 3.9, respectively. Age at diagnosis, year of diagnosis, tumor grade, and treatment modalities are prognostic factors for survival. The hazard ratios for patients who received surgery plus chemotherapy, surgery alone, and supportive care were 0.47 (95% CI 0.44–0.51), 1.22 (95% CI 1.1–1.36), and 3.23 (95% CI 3.01–3.46), respectively, by multivariable Cox regression analysis when using chemotherapy alone as a referent. Chemotherapy plus surgery may have a survival benefit for some selected gastric adenocarcinoma patients with metastatic disease.

RespireNet: A Deep Neural Network for Accurately Detecting Abnormal Lung Sounds in Limited Data Setting.

Auscultation of respiratory sounds is the primary tool for screening and diagnosing lung diseases. Automated analysis, coupled with digital stethoscopes, can play a crucial role in enabling tele-screening of fatal lung diseases. Deep neural networks (DNNs) have shown potential to solve such problems, and are an obvious choice. However, DNNs are data hungry, and the largest respiratory dataset ICBHI has only 6898 breathing cycles, which is quite small for training a satisfactory DNN model. In this work, RespireNet, we propose a simple CNN-based model, along with a suite of novel techniques- device specific fine-tuning, concatenation-based augmentation, blank region clipping, and smart padding-enabling us to efficiently use the small-sized dataset. We perform extensive evaluation on the ICBHI dataset, and improve upon the state-of-the-art results for 4-class classification by 2.2%.Code: https://github.com/microsoft/RespireNet.

The multivariate mixed Negative Binomial regression model with an application to insurance a posteriori ratemaking

This paper is concerned with introducing a family of multivariate mixed Negative Binomial regression models in the context of a posteriori ratemaking. The multivariate mixed Negative Binomial regression model can be considered as a candidate model for capturing overdispersion and positive dependencies in multi-dimensional claim count data settings, which all recent studies suggest are the norm when the ratemaking consists of pricing different types of claim counts arising from the same policy. For expository purposes, we consider the bivariate Negative Binomial-Gamma and Negative Binomial-Inverse Gaussian regression models. An Expectation-Maximization type algorithm is developed for maximum likelihood estimation of the parameters of the models for which the definition of a joint probability mass function in closed form is not feasible when the marginal means are modelled in terms of covariates. In order to illustrate the versatility of the proposed estimation procedure a numerical illustration is performed on motor insurance data on the number of claims from third party liability bodily injury and property damage. Finally, the a posteriori, or Bonus-Malus, premium rates resulting from the bivariate Negative Binomial-Gamma and Negative Binomial-Inverse Gaussian regression model are compared to those determined by the bivariate Negative Binomial and Poisson-Inverse Gaussian regression models.

An overview of AI and Machine Learning methods: motivations, concepts, and examples

Abstract What is artificial intelligence? Big Data? Machine learning? It may seem very difficult to define these terms and identify their use in real life, aside from the hype surrounding them over the past ten years. All of these terms are interconnected - sometimes they intersect, often they are included in each other. The main focus of this presentation is to present ‘Machine learning' methods, which can be defined as algorithms that learn patterns (a function) about a particular phenomenon (an output variable), based on examples (training data), often without any a priori on the shape of these patterns. Therefore, they are a broad family of methods that include traditional statistical methods such as linear regression models, but which often offer more flexibility in terms of hypothesis and complex data settings. Multiple frameworks exist, based on how many examples are available (size of the training data) and the kind of modeling application one wants to perform (e.g., classification, regression, clustering). Many different families of methods exist for each modeling application, which offers a tremendous amount of different algorithms, each having advantages and drawbacks. In practice, easy-to-use libraries were developed in popular programming languages such as Python and R, offering implementations of the majority of essential algorithms.

Node copying: A random graph model for effective graph sampling

There has been an increased interest in applying machine learning techniques on relational structured-data based on an observed graph. Often, this graph is not fully representative of the true relationship amongst nodes. In these settings, building a generative model conditioned on the observed graph allows to take the graph uncertainty into account. Various existing techniques either rely on restrictive assumptions, fail to preserve topological properties within the samples or are prohibitively expensive for larger graphs. In this work, we introduce the node copying model for constructing a distribution over graphs. Sampling of a random graph is carried out by replacing each node’s neighbors by those of a randomly sampled similar node. The sampled graphs preserve key characteristics of the graph structure without explicitly targeting them. Additionally, sampling from this model is extremely simple and scales linearly with the nodes. We show the usefulness of the copying model in three tasks. First, in node classification, a Bayesian formulation based on node copying achieves higher accuracy in sparse data settings. Second, we employ our proposed model to mitigate the effect of adversarial attacks on the graph topology. Last, incorporation of the model in a recommendation system setting improves recall over state-of-the-art methods.

Open Data Settings: A Conceptual Framework Explored Through the Map Room Project

In recent years, researchers have sought more effective ways of making data open, for purposes of accountability, engagement, and reuse. Often, such efforts focus on making existing data sets available to broad audiences. The expression data set itself suggests something discrete, complete, and easily transferable. But data are none of those things. In this paper, we argue that open data projects could benefit from a more contextual understanding of what open means. Instead of focusing on open data sets, researchers can seek to create and understand open data settings: contexts in which things of public significance can be presented as evidence. We share our experiences creating and analyzing open data settings for the Map Room Project, a research through design initiative that establishes local spaces for collaborative data exploration and mapping. Our contribution is to offer a conceptual framework through which researchers, as well as designers, might think about the openness of data settings. This framework emerged from our situational analysis of comparative empirical case studies. In Map Rooms, we found that open can mean accessible, inclusive, or indeterminate. Three practices of contextualization, which we call configuring, convening, and claim-making, shape these dimensions of openness by defining all of the following: where data can work, who is empowered to use them, and what can count as data.

Partial separability and functional graphical models for multivariate Gaussian processes

SummaryThe covariance structure of multivariate functional data can be highly complex, especially if the multivariate dimension is large, making extensions of statistical methods for standard multivariate data to the functional data setting challenging. For example, Gaussian graphical models have recently been extended to the setting of multivariate functional data by applying multivariate methods to the coefficients of truncated basis expansions. However, compared with multivariate data, a key difficulty is that the covariance operator is compact and thus not invertible. This paper addresses the general problem of covariance modelling for multivariate functional data, and functional Gaussian graphical models in particular. As a first step, a new notion of separability for the covariance operator of multivariate functional data is proposed, termed partial separability, leading to a novel Karhunen–Loève-type expansion for such data. Next, the partial separability structure is shown to be particularly useful in providing a well-defined functional Gaussian graphical model that can be identified with a sequence of finite-dimensional graphical models, each of identical fixed dimension. This motivates a simple and efficient estimation procedure through application of the joint graphical lasso. Empirical performance of the proposed method for graphical model estimation is assessed through simulation and analysis of functional brain connectivity during a motor task.

Bayesian inference of dependent kappa for binary ratings.

In medical and social science research, reliability of testing methods measured through inter- and intraobserver agreement is critical in disease diagnosis. Often comparison of agreement across multiple testing methods is sought in situations where testing is carried out on the same experimental units rendering the outcomes to be correlated. In this article, we first developed a Bayesian method for comparing dependent agreement measures under a grouped data setting. Simulation studies showed that the proposed methodology outperforms the competing methods in terms of power, while maintaining a decent type I error rate. We further developed a Bayesian joint model for comparing dependent agreement measures adjusting for subject and rater-level heterogeneity. Simulation studies indicate that our model outperforms a competing method that is used in this context. The developed methodology was implemented on a key measure on a dichotomous rating scale from a study with six raters evaluating three classification methods for chest radiographs for pneumoconiosis developed by the International Labor Office.