Data Science Challenges Research Articles

Unitarism vs. Individuality and a New Digital Agenda: The Power of Decentralized Web

Discussions around Covid-19 apps and models demonstrated that primary challenges for AI and data science focused on governance and ethics. Personal information was involved in building data sets. It was unclear how this information could be utilized in large scale models to provide predictions and insights while observing privacy requirements. Most people expected a lot from technology but were unwilling to sacrifice part of their privacy for building it. Conversely, regulators and policy makers require AI and data science practitioners to ensure optimal public health, national security while avoiding these privacy-related struggles. Their choices vary largely from country to country and are driven more by cultural aspects, and less by machine learning capabilities. The question is whether current ways to design technology and work with data sets are sustainable and lead to a good outcome for individuals and their communities. At the same time Covid-19 made it obvious that economies and societies cannot succeed without far-reaching digital policies, touching every aspect of how we provide and receive education, live, and work. Most regions, businesses and individuals struggled to benefit from competitive capabilities modern data technologies could bring. This opinion paper suggests how Germany and Europe can rethink their digital policy while recognizing the value of data, introducing Data IDs for consumers and businesses, committing to support innovation in decentralized data technologies, introducing concepts of Data Trusts and compulsory education around data starting from the early school age. Besides, it discusses advantages of data-tokens to shape a new ecosystem for decentralized data exchange. Furthermore, it emphasizes the necessity to develop and promote technologies to work with small data sets and handle data in compliance with privacy regulations, keeping in mind costs for the environment while bidding on big data and large-scale machine learning models. Finally, innovation as an integral part of any data scientist's job will be called for.

A Standards Organization for Open and FAIR Neuroscience: the International Neuroinformatics Coordinating Facility

There is great need for coordination around standards and best practices in neuroscience to support efforts to make neuroscience a data-centric discipline. Major brain initiatives launched around the world are poised to generate huge stores of neuroscience data. At the same time, neuroscience, like many domains in biomedicine, is confronting the issues of transparency, rigor, and reproducibility. Widely used, validated standards and best practices are key to addressing the challenges in both big and small data science, as they are essential for integrating diverse data and for developing a robust, effective, and sustainable infrastructure to support open and reproducible neuroscience. However, developing community standards and gaining their adoption is difficult. The current landscape is characterized both by a lack of robust, validated standards and a plethora of overlapping, underdeveloped, untested and underutilized standards and best practices. The International Neuroinformatics Coordinating Facility (INCF), an independent organization dedicated to promoting data sharing through the coordination of infrastructure and standards, has recently implemented a formal procedure for evaluating and endorsing community standards and best practices in support of the FAIR principles. By formally serving as a standards organization dedicated to open and FAIR neuroscience, INCF helps evaluate, promulgate, and coordinate standards and best practices across neuroscience. Here, we provide an overview of the process and discuss how neuroscience can benefit from having a dedicated standards body.

Assessment of the timeliness and robustness for predicting adult sepsis

SummarySepsis is a leading cause of death among inpatients at hospitals. However, with early detection, death rate can drop substantially. In this study, we present the top-performing algorithm for Sepsis II prediction in the DII National Data Science Challenge using the Cerner Health Facts data involving more than 100,000 adult patients. This large sample size allowed us to dissect the predictability by age-groups, race, genders, and care settings and up to 192 hr of sepsis onset. This large data collection also allowed us to conclude that the last six biometric records on average are informative to the prediction of sepsis. We identified biomarkers that are common across the treatment time and novel biomarkers that are uniquely presented for early prediction. The algorithms showed meaningful signals days ahead of sepsis onset, supporting the potential of reducing death rate by focusing on high-risk populations identified from heterogeneous data integration.

An interpretable deep-learning model for early prediction of sepsis in the emergency department

SummarySepsis is a life-threatening condition with high mortality rates and expensive treatment costs. Early prediction of sepsis improves survival in septic patients. In this paper, we report our top-performing method in the 2019 DII National Data Science Challenge to predict onset of sepsis 4 h before its diagnosis on electronic health records of over 100,000 unique patients in emergency departments. A long short-term memory (LSTM)-based model with event embedding and time encoding is leveraged to model clinical time series and boost prediction performance. Attention mechanism and global max pooling techniques are utilized to enable interpretation for the deep-learning model. Our model achieved an average area under the curve of 0.892 and was selected as one of the winners of the challenge for both prediction accuracy and clinical interpretability. This study paves the way for future intelligent clinical decision support, helping to deliver early, life-saving care to the bedside of septic patients.

Data Science Techniques, Assumptions, and Challenges in Alloy Clustering and Property Prediction

Data analytics methods have been increasingly applied to understanding materials chemistry, processing due to the manufacturing approach, and uni-axial and cyclic property relationships in the highly complex space of alloy design. There are several benefits to applying data analytics to this space, including the ability to manage non-linearities in the responses of the alloy attributes and the resulting mechanical properties. However, key difficulties in applying and understanding the results of data analytics include the often lack of reported assumptions and data processing steps necessary to improve interpretation and reproducibility in derived results. In this work, the methods used to generate clustering and correlation analyses for experimental 9% Cr ferritic-martensitic steel data were investigated and the resulting implications for mechanical property predictions were assessed. This work uses principal component analysis, partitioning around medoids, t-SNE, and k-means clustering to investigate trends in composition, processing and microstructure information with creep and tensile properties, building on work done previously using a smaller version of the same dataset. The initial assumptions, preprocessing steps and methods are investigated and outlined in order to depict the fine level of detail required to convey the steps taken to process data and produce analytical results. The variations in the resulting analyses are explored due to the influence of new and more varied data.

Predicting Drug Approvals: The Novartis Data Science and Artificial Intelligence Challenge

We describe a novel collaboration between academia and industry, an in-house data science and artificial intelligence challenge held by Novartis to develop machine learning models for predicting drug development outcomes, building upon research at MIT using data from Informa® as the starting point. Over 50 cross-functional teams from 25 Novartis offices around the world participated in the challenge. The domain expertise of these Novartis researchers was leveraged to create predictive models with greater sophistication, two teams developed models that outperformed the baseline MIT model through state-of-the-art machine learning algorithms and the use of newly incorporated features and data. In addition to validating the variables shown to be associated with drug approval in the earlier MIT study, the challenge also provided new insights into the drivers of drug development success and failure.

Acting with Inherently Uncertain Data: Practices of Data-Centric Knowing

Data-driven data science challenges our conceptualization of “data.” Significantly beyond capturing a given phenomenon, data, increasingly, are the phenomenon. Data may be iteratively manipulated algorithmically, undermining the “faithfulness” of data to any originating phenomenon. Crucially, data that are not “faithful” are inherently uncertain, as data risk becoming meaningless symbols. We empirically study how a community of commercially based geoscientists grapples with the phenomenon of offshore oil and gas reservoirs located kilometers below the seabed. The data available about these reservoirs are algorithmically manipulated sensor-based Internet of Things data. Our main contribution is the articulation of three patterns of work practices detailing how inherently uncertain data are woven into consequential work practices: (1) Accumulating is the cumulative process of supporting and triangulating one set of data with supplementary data sets; accumulating captures the conservative approach of backing up existing interpretations of the data. (2) Reframing describes the process where existing interpretations are contested by new data or models; reframing captures that there are limits to how far data can be manipulated. (3) Prospecting, or the cultivation of competing, incompatible data interpretations; while the former two patterns essentially attempt to regulate uncertainty, prospecting is about embracing it. Our concept of data-centric knowing is constituted by these three interwoven, ongoing practices.

Avoiding a Data Science Winter by Keeping the Expectations Low

In the paper we present and discuss some aspects related to what we consider as some of the most important corporate challenges of Data Science, AI and machine learning regarding both human talents and business. We examine the case of a discussion that took place over Quora and in particular we focus on an answer we have selected as indicative of a potentially threatening situation for the sustainable development of the data science, AI and machine learning disciplines as well as the growth of the respective demand and supply sides and the corresponding ecosystem these form. We then make an attempt to examine the setting by means of analyzing the case, using as our guide the provided narrative.

Merlin—A unified modeling framework for data analysis and methods development in Stata

The challenges in statistics and data science are rapidly growing because access to a multitude of data types continues to increase, as well as the sheer quantity of data. Analysts are now presented with multivariate data, sometimes measured repeatedly, and often requiring the ability to model nonlinear relationships and hierarchical structures. In this article, I present the merlin command, which attempts to provide an extremely general framework for data analysis. From simple settings such as fitting a linear regression model or a Weibull survival model to more complex settings such as fitting a three-level logistic mixed-effects model or a multivariate joint model of multiple longitudinal outcomes (of different types) and a recurrent event and survival with nonlinear effects, merlin can fit them all. I will take a single dataset and attempt to show you the full range of capabilities of merlin and discuss some future directions for the implementation in Stata.

Early-Career View on Data Science Challenges: Responsibility, Rigor, and Accessibility

Early-Career View on Data Science Challenges: Responsibility, Rigor, and Accessibility

The AirSensor open-source R-package and DataViewer web application for interpreting community data collected by low-cost sensor networks

While large-scale low-cost sensor networks are now recording air pollutant concentrations at finer spatial and temporal scales than previously measured, the large environmental data sets generated by these sensor networks can become overwhelming when considering the scientific skills required to analyze the data and generate interpretable results. This paper summarizes the development of an open-source R package (AirSensor) and interactive web application (DataViewer) designed to address the environmental data science challenges of visualizing and understanding local air quality conditions with community networks of low-cost air quality sensors. AirSensor allows users to access historical data, add spatial metadata, and create maps and plots for viewing community monitoring data. The DataViewer application was developed to incorporate the functionality and plotting functions of the R package into a user-friendly web experience that would serve as the primary source for data communication for community-based organizations and citizen scientists.

Distributed linguistic representations in decision making: Taxonomy, key elements and applications, and challenges in data science and explainable artificial intelligence

Distributed linguistic representations are powerful tools for modelling the uncertainty and complexity of preference information in linguistic decision making. To provide a comprehensive perspective on the development of distributed linguistic representations in decision making, we present the taxonomy of existing distributed linguistic representations. Then, we review the key elements and applications of distributed linguistic information processing in decision making, including the distance measurement, aggregation methods, distributed linguistic preference relations, and distributed linguistic multiple attribute decision making models. Next, we provide a discussion on ongoing challenges and future research directions from the perspective of data science and explainable artificial intelligence.

What Is Your List of 10 Challenges in Data Science?

What Is Your List of 10 Challenges in Data Science?

Biomedical Data Science and Informatics Challenges to Implementing Pharmacogenomics with Electronic Health Records

Pharmacogenomic information must be incorporated into electronic health records (EHRs) with clinical decision support in order to fully realize its potential to improve drug therapy. Supported by various clinical knowledge resources, pharmacogenomic workflows have been implemented in several healthcare systems. Little standardization exists across these efforts, however, which limits scalability both within and across clinical sites. Limitations in information standards, knowledge management, and the capabilities of modern EHRs remain challenges for the widespread use of pharmacogenomics in the clinic, but ongoing efforts are addressing these challenges. Although much work remains to use pharmacogenomic information more effectively within clinical systems, the experiences of pioneering sites and lessons learned from those programs may be instructive for other clinical areas beyond genomics. We present a vision of what can be achieved as informatics and data science converge to enable further adoption of pharmacogenomics in the clinic.

Codification Challenges for Data Science in Construction

AbstractNew forms of data science, including machine learning and data analytics, are enabled by machine-readable information but are not widely deployed in construction. A qualitative study of inf...

A Network-Based Indicator of Travelers Performativity on Instagram

The spread of Internet and online social media has created a huge amount of data able to provide new insights to researchers in different disciplinary fields, but it also presents new challenges for data science. Data arising from online social networks can be naturally coded as relational data in affiliation and adjacency matrices, then analyzed with social network analysis. In this study, we apply an interdisciplinary approach (based on automatic visual content analysis, social network analysis, and exploratory statistical techniques) to define and derive a suitable indicator for characterizing places, along with the online activities of travelers, in terms of sharing images. We envisage a novel storytelling perspective where stories are related to places and the narrative activity is realized through posting images. Specifically, we use data extracted from an online social network (i.e., Instagram) to identify travelers’ paths among sites of interests. Starting from a large collection of pictures geolocalized in a pre-specified set of locations (i.e., five locations in the Campania region of Italy during the 2018 Christmas season), we use automatic alternative text to produce an ex-post taxonomy of images on the most recurrent themes emerging from pictures posted on Instagram. Quantitative measures defined on the co-occurrence of locations and the emerging themes are used to build a statistical indicator able to characterize paths among different locations as narrated from travelers’ posts. The proposed analysis, presented and discussed along with real data, can be useful for stakeholders interested in the fields of policy-making, communication design, and territory profiling strategies.

Emerging Priorities for Microbiome Research.

Microbiome research has increased dramatically in recent years, driven by advances in technology and significant reductions in the cost of analysis. Such research has unlocked a wealth of data, which has yielded tremendous insight into the nature of the microbial communities, including their interactions and effects, both within a host and in an external environment as part of an ecological community. Understanding the role of microbiota, including their dynamic interactions with their hosts and other microbes, can enable the engineering of new diagnostic techniques and interventional strategies that can be used in a diverse spectrum of fields, spanning from ecology and agriculture to medicine and from forensics to exobiology. From June 19–23 in 2017, the NIH and NSF jointly held an Innovation Lab on Quantitative Approaches to Biomedical Data Science Challenges in our Understanding of the Microbiome. This review is inspired by some of the topics that arose as priority areas from this unique, interactive workshop. The goal of this review is to summarize the Innovation Lab’s findings by introducing the reader to emerging challenges, exciting potential, and current directions in microbiome research. The review is broken into five key topic areas: (1) interactions between microbes and the human body, (2) evolution and ecology of microbes, including the role played by the environment and microbe-microbe interactions, (3) analytical and mathematical methods currently used in microbiome research, (4) leveraging knowledge of microbial composition and interactions to develop engineering solutions, and (5) interventional approaches and engineered microbiota that may be enabled by selectively altering microbial composition. As such, this review seeks to arm the reader with a broad understanding of the priorities and challenges in microbiome research today and provide inspiration for future investigation and multi-disciplinary collaboration.

K-Means-based isolation forest

The task of anomaly detection in data is one of the main challenges in data science because of the wide plethora of applications and despite a spectrum of available methods. Unfortunately, many of anomaly detection schemes are still imperfect i.e., they are not effective enough or act in a non-intuitive way or they are focused on a specific type of data. In this study, the classical method of Isolation Forest is thoroughly analyzed and augmented by bringing an innovative approach. This is k-Means-Based Isolation Forest that allows to build a search tree based on many branches in contrast to the only two considered in the original method. k-Means clustering is used to predict the number of divisions on each decision tree node. As supported through experimental studies, the proposed method works effectively for data coming from various application areas including intermodal transport and geographical, spatio-temporal data. In addition, it enables a user to intuitively determine the anomaly score for an individual record of the analyzed dataset. The advantage of the proposed method is that it is able to fit the data at the step of decision tree building. Moreover, it returns more intuitively appealing anomaly score values.

Open Science principles for accelerating trait-based science across the Tree of Life.

Synthesizing trait observations and knowledge across the Tree of Life remains a grand challenge for biodiversity science. Species traits are widely used in ecological and evolutionary science, and new data and methods have proliferated rapidly. Yet accessing and integrating disparate data sources remains a considerable challenge, slowing progress toward a global synthesis to integrate trait data across organisms. Trait science needs a vision for achieving global integration across all organisms. Here, we outline how the adoption of key Open Science principles-open data, open source and open methods-is transforming trait science, increasing transparency, democratizing access and accelerating global synthesis. To enhance widespread adoption of these principles, we introduce the Open Traits Network (OTN), a global, decentralized community welcoming all researchers and institutions pursuing the collaborative goal of standardizing and integrating trait data across organisms. We demonstrate how adherence to Open Science principles is key to the OTN community and outline five activities that can accelerate the synthesis of trait data across the Tree of Life, thereby facilitating rapid advances to address scientific inquiries and environmental issues. Lessons learned along the path to a global synthesis of trait data will provide a framework for addressing similarly complex data science and informatics challenges.

Crowd-Sourcing for Data Science and Quantifiable Challenges: Optimal Contest Design

We study the optimal design of a crowd-sourcing contest in settings where the output (from the contestants) is quantifiable -- for example, a data science challenge. This setting is in contrast to settings where the output is only qualitative and cannot be quantified in an objective manner -- for example, when the goal of the contest is to design a logo. The rapidly growing literature on the design of crowd-sourcing contests focuses largely on ordinal contests -- these are contests where contestants' outputs are ranked by the organizer and awards are based on the relative ranks. Such contests are ideally suited for the latter setting, where output is qualitative. For our setting (quantitative output), it is possible to design contests where awards are based on the actual outputs and not on their ranking alone -- thus, our space of contest designs includes ordinal contests but is significantly larger. We derive an easy-to-implement contest design for this setting and establish its optimality.