Data Science Challenges Research Articles

JUNGFRAU detector for brighter x-ray sources: Solutions for IT and data science challenges in macromolecular crystallography.

In this paper, we present a data workflow developed to operate the adJUstiNg Gain detector FoR the Aramis User station (JUNGFRAU) adaptive gain charge integrating pixel-array detectors at macromolecular crystallography beamlines. We summarize current achievements for operating at 9 GB/s data-rate a JUNGFRAU with 4 Mpixel at 1.1 kHz frame-rate and preparations to operate at 46 GB/s data-rate a JUNGFRAU with 10 Mpixel at 2.2 kHz in the future. In this context, we highlight the challenges for computer architecture and how these challenges can be addressed with innovative hardware including IBM POWER9 servers and field-programmable gate arrays. We discuss also data science challenges, showing the effect of rounding and lossy compression schemes on the MX JUNGFRAU detector images.

Protein Interaction Networks: Functional and Statistical Approaches.

The evolution of next-generation sequencing and high-throughput technologies has created new opportunities and challenges in data science. Currently, a classic proteomics analysis can be complemented by going a step beyond the individual analysis of the proteome by using integrative approaches. These integrations can be focused either on inferring relationships among proteins themselves, with other molecular levels, phenotype, or even environmental data, giving the researcher new tools to extract and determine the most relevant information in biological terms. Furthermore, it is also important the employ of visualization methods that allow a correct and deep interpretation of data.To carry out these analyses, several bioinformatics and biostatistical tools are required. In this chapter, different workflows that enable the creation of interaction networks are proposed. Resulting networks reduce the complexity of original datasets, depicting complex statistical relationships (through PLS analysis and variants), functional networks (STRING, shinyGO), and a combination of both approaches. Recently developed methods for integrating different omics levels, such as coinertial analyses or DIABLO, are also described. Finally, the use of Cytoscape or Gephi was described for the representation and mining of the different networks.This approach constitutes a new way of acquiring a deeper knowledge of the function of proteins, such as the search for specific connections of each group to identify differentially connected modules, which may reflect involved protein complexes and key pathways.

GO FAIR Brazil: A Challenge for Brazilian Data Science

The FAIR principles, an acronym for Findable, Accessible, Interoperable and Reusable, are recognised worldwide as key elements for good practice in all data management processes. To understand how the Brazilian scientific community is adhering to these principles, this article reports Brazilian adherence to the GO FAIR initiative through the creation of the GO FAIR Brazil Office and the manner in which they create their implementation networks. To contextualise this understanding, we provide a brief presentation of open data policies in Brazilian research and government, and finally, we describe a model that has been adopted for the GO FAIR Brazil implementation networks. The Brazilian Institute of Information in Science and Technology is responsible for the GO FAIR Brazil Office, which operates in all fields of knowledge and supports thematic implementation networks. Today, GO FAIR Brazil-Health is the first active implementation network in operation, which works in all health domains, serving as a model for other fields like agriculture, nuclear energy, and digital humanities, which are in the process of adherence negotiation. This report demonstrates the strong interest and effort from the Brazilian scientific communities in implementing the FAIR principles in their research data management practices.

Tackling Overfitting in Boosting for Noisy Healthcare Data

Analyzing healthcare data poses several challenges including the limited number of samples, missing measurements, noisy labels, and heterogeneous data types. Tree-based boosting is well-suited for modeling such data as it is insensitive to data types and missingness. Moreover, Stochastic Gradient TreeBoost is often found in many winning solutions in public data science challenges. Unfortunately, the best performance requires extensive hyperparameter tuning and can be prone to overfitting. We propose PaloBoost, a Stochastic Gradient TreeBoost model that uses novel regularization techniques to guard against overfitting and is robust to hyperparameter settings. PaloBoost uses the out-of-bag samples to perform gradient-aware pruning and estimate adaptive learning rates. Unlike other Stochastic Gradient TreeBoost models that use the out-of-bag samples to estimate test errors, PaloBoost treats the samples as a second batch of training samples to prune the trees and adjust the learning rates. As a result, PaloBoost can dynamically adjust tree depths and learning rates to achieve faster learning at the start and slower learning as the algorithm converges. Experimental results on four datasets demonstrate that PaloBoost is robust to overfitting and is less sensitive to the hyperparameters.

Data Science of the Natural Environment: A Research Roadmap

Data science is the science of extracting meaning from potentially complex data. This is a fast moving field, drawing principles and techniques from a number of different disciplinary areas including computer science, statistics and complexity science. Data science is having a profound impact on a number of areas including commerce, health and smart cities. This paper argues that data science can have an equal if not greater impact in the area of earth and environmental sciences, offering a rich tapestry of new techniques to support both a deeper understanding of the natural environment in all its complexities, as well as the development of well-founded mitigation and adaptation strategies in the face of climate change. The paper argues that data science for the natural environment brings about new challenges for data science, particularly around complexity, spatial and temporal reasoning, and managing uncertainty. The paper also describes a case study in environmental data science which offers up insights into the promise of the area. The paper concludes with a research roadmap highlighting ten top challenges of environmental data science and also an invitation to become part of an international community working collaboratively on these problems.

On the Role of Bioinformatics and Data Science in Industrial Microbiome Applications.

Advances in sequencing and computational biology have drastically increased our capability to explore the taxonomic and functional compositions of microbial communities that play crucial roles in industrial processes. Correspondingly, commercial interest has risen for applications where microbial communities make important contributions. These include food production, probiotics, cosmetics, and enzyme discovery. Other commercial applications include software that takes the user’s gut microbiome data as one of its inputs and outputs evidence-based, automated, and personalized diet recommendations for balanced blood sugar levels. These applications pose several bioinformatic and data science challenges that range from requiring strain-level resolution in community profiles to the integration of large datasets for predictive machine learning purposes. In this perspective, we provide our insights on such challenges by touching upon several industrial areas, and briefly discuss advances and future directions of bioinformatics and data science in microbiome research.

Particulate Matter Matters—The Data Science Challenge @ BTW 2019

For the second time, the Data Science Challenge took place as part of the 18th symposium “Database Systems for Business, Technology and Web” (BTW) of the Gesellschaft fur Informatik (GI). The Challenge was organized by the University of Rostock and sponsored by IBM and SAP. This year, the integration, analysis and visualization around the topic of particulate matter pollution was the focus of the challenge. After a preselection round, the accepted participants had one month to adapt their developed approach to a substantiated problem, the real challenge. The final presentation took place at BTW 2019 in front of the prize jury and the attending audience. In this article, we give a brief overview of the schedule and the organization of the Data Science Challenge. In addition, the problem to be solved and its solution will be presented by the participants.

Fourier could be a data scientist: From graph Fourier transform to signal processing on graphs

Abstract The legacy of Joseph Fourier in science is vast, especially thanks to the essential tool that the Fourier transform is. The flexibility of this analysis, its computational efficiency and the physical interpretation it offers makes it a cornerstone in many scientific domains. With the explosion of digital data, both in quantity and diversity, the generalization of the tools based on Fourier transform is mandatory. In data science, new problems arose for the processing of irregular data such as social networks, biological networks or other data on networks. Graph signal processing is a promising approach to deal with those. The present text is an overview of the state of the art in graph signal processing, focusing on how to define a Fourier transform for data on graphs, how to interpret it and how to use it to process such data. It closes showing some examples of use. Along the way, the review reveals how Fourier's work remains modern and universal, and how his concepts, coming from physics and blended with mathematics, computer science, and signal processing, play a key role in answering the modern challenges in data science.

Geometric Characterization of Data Sets with Unique Reduced Gröbner Bases.

Model selection based on experimental data is an important challenge in biological data science. Particularly when collecting data is expensive or time-consuming, as it is often the case with clinical trial and biomolecular experiments, the problem of selecting information-rich data becomes crucial for creating relevant models. We identify geometric properties of input data that result in an unique algebraic model, and we show that if the data form a staircase, or a so-called linear shift of a staircase, the ideal of the points has a unique reduced Gröbner basis and thus corresponds to a unique model. We use linear shifts to partition data into equivalence classes with the same basis. We demonstrate the utility of the results by applying them to a Boolean model of the well-studied lac operon in E. coli.

Edge2vec: Representation learning using edge semantics for biomedical knowledge discovery

BackgroundRepresentation learning provides new and powerful graph analytical approaches and tools for the highly valued data science challenge of mining knowledge graphs. Since previous graph analytical methods have mostly focused on homogeneous graphs, an important current challenge is extending this methodology for richly heterogeneous graphs and knowledge domains. The biomedical sciences are such a domain, reflecting the complexity of biology, with entities such as genes, proteins, drugs, diseases, and phenotypes, and relationships such as gene co-expression, biochemical regulation, and biomolecular inhibition or activation. Therefore, the semantics of edges and nodes are critical for representation learning and knowledge discovery in real world biomedical problems.ResultsIn this paper, we propose the edge2vec model, which represents graphs considering edge semantics. An edge-type transition matrix is trained by an Expectation-Maximization approach, and a stochastic gradient descent model is employed to learn node embedding on a heterogeneous graph via the trained transition matrix. edge2vec is validated on three biomedical domain tasks: biomedical entity classification, compound-gene bioactivity prediction, and biomedical information retrieval. Results show that by considering edge-types into node embedding learning in heterogeneous graphs, edge2vec significantly outperforms state-of-the-art models on all three tasks.ConclusionsWe propose this method for its added value relative to existing graph analytical methodology, and in the real world context of biomedical knowledge discovery applicability.

A data science challenge for converting airborne remote sensing data into ecological information.

Ecology has reached the point where data science competitions, in which multiple groups solve the same problem using the same data by different methods, will be productive for advancing quantitative methods for tasks such as species identification from remote sensing images. We ran a competition to help improve three tasks that are central to converting images into information on individual trees: (1) crown segmentation, for identifying the location and size of individual trees; (2) alignment, to match ground truthed trees with remote sensing; and (3) species classification of individual trees. Six teams (composed of 16 individual participants) submitted predictions for one or more tasks. The crown segmentation task proved to be the most challenging, with the highest-performing algorithm yielding only 34% overlap between remotely sensed crowns and the ground truthed trees. However, most algorithms performed better on large trees. For the alignment task, an algorithm based on minimizing the difference, in terms of both position and tree size, between ground truthed and remotely sensed crowns yielded a perfect alignment. In hindsight, this task was over simplified by only including targeted trees instead of all possible remotely sensed crowns. Several algorithms performed well for species classification, with the highest-performing algorithm correctly classifying 92% of individuals and performing well on both common and rare species. Comparisons of results across algorithms provided a number of insights for improving the overall accuracy in extracting ecological information from remote sensing. Our experience suggests that this kind of competition can benefit methods development in ecology and biology more broadly.

Individual tree crown delineation and tree species classification with hyperspectral and LiDAR data.

An international data science challenge, called National Ecological Observatory Network—National Institute of Standards and Technology data science evaluation, was set up in autumn 2017 with the goal to improve the use of remote sensing data in ecological applications. The competition was divided into three tasks: (1) individual tree crown (ITC) delineation, for identifying the location and size of individual trees; (2) alignment between field surveyed trees and ITCs delineated on remote sensing data; and (3) tree species classification. In this paper, the methods and results of team Fondazione Edmund Mach (FEM) are presented. The ITC delineation (Task 1 of the challenge) was done using a region growing method applied to a near-infrared band of the hyperspectral images. The optimization of the parameters of the delineation algorithm was done in a supervised way on the basis of the Jaccard score using the training set provided by the organizers. The alignment (Task 2) between the delineated ITCs and the field surveyed trees was done using the Euclidean distance among the position, the height, and the crown radius of the ITCs and the field surveyed trees. The classification (Task 3) was performed using a support vector machine classifier applied to a selection of the hyperspectral bands and the canopy height model. The selection of the bands was done using the sequential forward floating selection method and the Jeffries Matusita distance. The results of the three tasks were very promising: team FEM ranked first in the data science competition in Task 1 and 2, and second in Task 3. The Jaccard score of the delineated crowns was 0.3402, and the results showed that the proposed approach delineated both small and large crowns. The alignment was correctly done for all the test samples. The classification results were good (overall accuracy of 88.1%, kappa accuracy of 75.7%, and mean class accuracy of 61.5%), although the accuracy was biased toward the most represented species.

Beyond the data management plan: Expanding roles for librarians in data science and open science

ABSTRACTSupport for research data management (RDM) has become a popular focus for library services, and many training opportunities in RDM exist for librarians. However, fewer opportunities exist for librarians to develop the more advanced skills and expertise necessary to support data science and open science. To better understand how to develop a library workforce prepared to develop services in the emerging areas of data science and open science, the National Library of Medicine convened a workshop of 15 librarians and information professionals with a range of expertise in data and open science. This session will present the workshop's recommendations and use participatory facilitation methods to engage attendees in conversation about how to prepare information professionals to meet the evolving challenges of data science and open science.

In Search of Disruptive Ideas - Outlier Detection Techniques in Crowdsourcing Innovation Platforms

The key challenge for data science in open innovation web systems is to find best ideas among thousands of community submissions. To date, this has been done with metrics reflecting enterprise needs or community preferences. This article proposes to look in a different direction: inspired by theoretical studies on disruptive innovation, we frame the problem of valuable ideas as those rarely taken up by masses or organisations yet having potential to change industries. Our aim is to find technological means for automatic detection of such innovations to aid decision making. Following past findings from business sciences on nature of disruptive innovations, the article presents a comparative study of multiple outlier detection algorithms applied to two real-world datasets containing textual descriptions of ideas for different industries. Obtained results demonstrate capability of outlier detection and show k-NN algorithm with TF-IDF and cosine distance to be the best candidate for the task.

Applying deep learning to right whale photo identification.

Photo identification is an important tool for estimating abundance and monitoring population trends over time. However, manually matching photographs to known individuals is time‐consuming. Motivated by recent developments in image recognition, we hosted a data science challenge on the crowdsourcing platform Kaggle to automate the identification of endangered North Atlantic right whales (Eubalaena glacialis). The winning solution automatically identified individual whales with 87% accuracy with a series of convolutional neural networks to identify the region of interest on an image, rotate, crop, and create standardized photographs of uniform size and orientation and then identify the correct individual whale from these passport‐like photographs. Recent advances in deep learning coupled with this fully automated workflow have yielded impressive results and have the potential to revolutionize traditional methods for the collection of data on the abundance and distribution of wild populations. Presenting these results to a broad audience should further bridge the gap between the data science and conservation science communities.

Population Data Science: The science of data about people

IntroductionSocietal and individual benefits of data-intensive science are substantial but raise challenges of balancing individual privacy and public good, while building appropriate governance and socio-technical systems to support data-intensive science. We set out to define a new field of inquiry to move collective interests forward.
 Objectives and ApproachOur objectives were: 1. To create a concise definition of the emerging field of Population Data Science; 2. To highlight the characteristics and challenges of Population Data Science; 3. To differentiate Population Data Science from existing fields of data science and informatics; and 4. To discuss the implications and future opportunities for Population Data Science. Objectives 1 and 2 were met largely through International Population Data Linkage Network (IPDLN) member engagement, Objective 3 was evaluated via literature review, and Objective 4 was achieved through iterative and collective work on a peer-reviewed position paper.
 ResultsWe define Population Data Science succinctly as the science of data about people. It is related to, but distinct from, the fields of data science and informatics. A broader definition includes four characteristics of: i) data use for positive impact on individuals and populations; ii) bringing together and analyzing data from multiple sources; iii) identifying population-level insights; and iv) developing safe, privacy-sensitive and ethical infrastructure to support research. One implication of these characteristics is that few individuals or organisations possess all of the requisite knowledge and skills comprising Population Data Science, so this is by nature a multi-disciplinary “team science” field. There is a need to advance various aspects of science, such as data linkage technology, various forms of analytics, and methods of public engagement.
 Conclusion/ImplicationsThese implications are the beginnings of a research agenda for Population Data Science, which if approached as a collective field, will catalyze significant advances in our understanding of society, health, and human behavior and increase the impact of our research.

Environmental Data Science

Environmental data are growing in complexity, size, and resolution. Addressing the types of large, multidisciplinary problems faced by today's environmental scientists requires the ability to leverage available data and information to inform decision making. Successfully synthesizing heterogeneous data from multiple sources to support holistic analyses and extraction of new knowledge requires application of Data Science. In this paper, we present the origins and a brief history of Data Science. We revisit prior efforts to define Data Science and provide a more modern, working definition. We describe the new professional profile of a data scientist and new and emerging applications of Data Science within Environmental Sciences. We conclude with a discussion of current challenges for Environmental Data Science and suggest a path forward.

A Position Statement on Population Data Science: The Science of Data about People.

Information is increasingly digital, creating opportunities to respond to pressing issues about human populations using linked datasets that are large, complex, and diverse. The potential social and individual benefits that can come from data-intensive science are large, but raise challenges of balancing individual privacy and the public good, building appropriate socio-technical systems to support data-intensive science, and determining whether defining a new field of inquiry might help move those collective interests and activities forward. A combination of expert engagement, literature review, and iterative conversations led to our conclusion that defining the field of Population Data Science (challenge 3) will help address the other two challenges as well. We define Population Data Science succinctly as the science of data about people and note that it is related to but distinct from the fields of data science and informatics. A broader definition names four characteristics of: data use for positive impact on citizens and society; bringing together and analyzing data from multiple sources; finding population-level insights; and developing safe, privacy-sensitive and ethical infrastructure to support research. One implication of these characteristics is that few people possess all of the requisite knowledge and skills of Population Data Science, so this is by nature a multi-disciplinary field. Other implications include the need to advance various aspects of science, such as data linkage technology, various forms of analytics, and methods of public engagement. These implications are the beginnings of a research agenda for Population Data Science, which if approached as a collective field, can catalyze significant advances in our understanding of trends in society, health, and human behavior.

Environmental data stream mining through a case-based stochastic learning approach

Environmental data stream mining is an open challenge for Data Science. Common methods used are static because they analyze a static set of data, and provide static data-driven models. Environmental systems are dynamic and generate a continuous data stream. Dynamic methods coping with the temporal nature of data must be provided in Data Science. Our proposal is to model each environmental information unit, timely generated, as a new case/experience in a Case-Based Reasoning (CBR) system. This contribution aims to incrementally build and manage a Dynamic Adaptive Case Library (DACL). In this paper, a stochastic method for the learning of new cases and management of prototypes to create and manage the DACL in an incremental way is introduced. This stochastic method works with two main moments. An evaluation of the method has been carried using a data stream of air quality of the city of Obregon, Sonora. México, with good results. In addition, other datasets have been mined to ensure the generality of the approach.

Prof. Fionn Murtagh: clinical and all medical practice, and the health context, the major benefits and some challenges of Big Data science

Fionn Murtagh ( Figure 1 ) is Professor of Data Science and was Professor of Computer Science, including Department Head, in many universities.