ArticlePDF Available

The Process of Analyzing Data is the Emergent Feature of Data Science

February 2016
Frontiers in Genetics 7(12)

DOI:10.3389/fgene.2016.00012

License
CC BY 4.0

Authors:

Frank Emmert-Streib

Tampere University

Salissou Moutari

Queen's University Belfast

Matthias Dehmer

TU Wien

In recent years the term “data science” gained considerable attention worldwide. In a A Very Short History Of Data Science by Press (2013), the first appearance of the term is ascribed to Peter Naur in 1974 (Concise Survey of Computer Methods). Regardless who used the term first and in what context it has been used, we think that data science is a good term to indicate that data are the focus of scientific research. This is in analogy to computer science, where the first department of computer science in the USA had been established in 1962 at Purdue University, at a time when the first electronic computers became available and it was still not clear enough what computers can do, one created therefore a new field where the computer was the focus of the study. In this paper, we want to address a couple of questions in order to demystify the meaning and the goals of data science in general.

| Schematic visualization of the constituting parts of data science and other disciplines in terms of the involvement of (1) domain knowledge, (2) statistics/mathematics, and (3) computer science.

…

Figures - available via license: Creative Commons Attribution 4.0 International

Content may be subject to copyright.

Available via license: CC BY 4.0

Content may be subject to copyright.

OPINION

published: 09 February 2016

doi: 10.3389/fgene.2016.00012

Frontiers in Genetics | www.frontiersin.org 1February 2016 | Volume 7 | Article 12

Edited by:

Celia M.T. Greenwood,

Lady Davis Institute, Canada

Reviewed by:

Robert Nadon,

McGill University, Canada

*Correspondence:

Frank Emmert-Streib

v@bio-complexity.com

Specialty section:

This article was submitted to

Statistical Genetics and Methodology,

a section of the journal

Frontiers in Genetics

Received: 25 October 2015

Accepted: 25 January 2016

Published: 09 February 2016

Citation:

Emmert-Streib F, Moutari S and

Dehmer M (2016) The Process of

Analyzing Data is the Emergent

Feature of Data Science.

Front. Genet. 7:12.

doi: 10.3389/fgene.2016.00012

The Process of Analyzing Data is the

Emergent Feature of Data Science

Frank Emmert-Streib 1*, Salissou Moutari 2and Matthias Dehmer 3

1Computational Medicine and Statistical Learning Laboratory, Department of Signal Processing, Tampere University of

Technology, Tampere, Finland, 2Centre for Statistical Science and Operational Research, School of Mathematics and

Physics, Queen’s University Belfast, Belfast, UK, 3Department of Computer Science, Institute for Theoretical Informatics,

Mathematics and Operations Research, Universität der Bundeswehr München, Neubiberg, Germany

Keywords: data science, computational biology, statistics, big data, high-throughput data

In recent years the term “data science” gained considerable attention worldwide. In a A Very Short

History Of Data Science by Press (2013), the ﬁrst appearance of the term is ascribed to Peter Naur

in 1974 (Concise Survey of Computer Methods). Regardless who used the term ﬁrst and in what

context it has been used, we think that data science is a good term to indicate that data are the

focus of scientiﬁc research. This is in analogy to computer science, where the ﬁrst department of

computer science in the USA had been established in 1962 at Purdue University, at a time when the

ﬁrst electronic computers became available and it was still not clear enough what computers can

do, one created therefore a new ﬁeld where the computer was the focus of the study. In this paper,

we want to address a couple of questions in order to demystify the meaning and the goals of data

science in general.

The ﬁrst question that comes to mind when hearing there is a new ﬁeld is, what makes such

existing ﬁeld diﬀerent from others? For this purpose Drew Conway created the data science Venn

diagram (Conway, 2003) that is helpful in this discussion. In Figure 1, we show a modiﬁed version

of the original diagram as an Efron-triangle (Efron, 2003), which includes metric information.

The important point to realize is that data science is not an entirely new ﬁeld in the sense that it

deals with problems outside of any other ﬁeld. Instead, the new contribution is its composition,

consisting of at least three major ﬁelds, or dimensions, namely (1) domain knowledge, (2)

statistics/mathematics, and (3) computer science. Here “domain knowledge” corresponds to a ﬁeld

that generates the data, e.g., biology, economics, ﬁnance, medicine, sociology, psychology etc. The

position of a particular ﬁeld in Figure 1, respectively the distances to the three corners of the

triangle, i.e., (d1,d2,d3), provide information about the contribution of the three major ﬁelds,

which can be seen as proportions or weights (see the examples in Figure 1). Overall, data science

emerges at the intersection of these three ﬁelds whereas the term emerges is important because there

is more than just “adding” the three parts together.

Before we come back to the discussion of the emergent aspect of data science let us give

some speciﬁc examples for existing ﬁelds in the light of data science in particular application

domains. Scientiﬁc ﬁelds with a long history analyzing data by means of statistical methods are

biostatistics and applied statistics. However, the computer science component in these ﬁelds is not

very noticeable despite the fact that also algorithmic methods are used, e.g., via software package

like SPSS (Statistical Package for the Social Sciences) or SAS (Statistical Analysis System). However,

such software packages are not fully ﬂexible programming languages but provide only a rather

limited set of statistical and visualization functions that can be applied to a data set, usually, via a

graphical user interface (GUI). Also, the preprocessing of the data sets themselves is diﬃcult within

the provided capabilities of such packages because of the lack of basic data manipulation functions.

A more recent ﬁeld that evolved from elements in biostatistics and bioinformatics is computational

biology. Depending on its deﬁnition, which diﬀers somewhat between the US and Europe, in

general, computational biology is an example for a data science with a speciﬁc focus on biological

Emmert-Streib et al. Data Science

FIGURE 1 | Schematic visualization of the constituting parts of data science and other disciplines in terms of the involvement of (1) domain

knowledge, (2) statistics/mathematics, and (3) computer science.

and biomedical data. This is especially true since the completion

of the Human Genome Project has led to a series of new

and aﬀordable high-throughput technologies that allow the

generating of a variety of diﬀerent types of ’omics data

(Quackenbush, 2011). Also, initiatives like The Cancer Genome

Atlas (TCGA) (The Cancer Genome Atlas Research Network,

2008) making the results of such large-scale experiments publicly

available in the form of databases contributed considerably to the

establishment of computational biology as a data science, because

without such data availability, there would not be data science at

all.

Maybe the best non-biological example of a problem that is

purely based on data is the Stock Market. Here the prices of shares

and stocks are continuously determined by the demand and

supply level of electronic transactions enabled by the diﬀerent

markets. For instance, the goal of day traders is to recognize

stable patterns within ordinary stochastic variations of price

movements to forecast future prices reliably. Due to the fact that

a typical time scale of a buying and selling cycle is within (a

fraction of) a trading day these prices are usually beyond real

value diﬀerences, e.g., of productivity changes of the companies

themselves corresponding to the traded shares, but are more

an expression of the expectations of the shareholders. For

this reason, economic knowledge about balance sheets, income

statements and cash ﬂow statements are not suﬃcient for making

informed trading decisions because the chart patterns need to be

analyzed themselves by means of statistical algorithms.

From Figure 1, one might get the feeling that a data scientist

is expected to have every skill of all three major ﬁelds. This

is not true, but merely an impression of the two-dimensional

projection of a multidimensional problem. For instance, it would

not be expected from a data scientist to prove theorems about

the convergence of a learning algorithm (Vapnik, 1995). This

would be within the skill set of a (mathematical) statistician

or a statistical learning theoretician. Also, conducting wet lab

experiments leading to the generation of the data itself is

outside the skill set. That means a data scientist needs to have

interdisciplinary skills from a couple of diﬀerent disciplines

but does not need to possess a complete skill set from all

of these ﬁelds. For this reason, these core disciplines do not

become redundant or obsolete, but will still make important

contributions beyond data science.

Importantly, there is an emergent component to data science

that cannot be explained by the linear summation of its

constituting elements described above. In our opinion this

emergent component is the dynamical aspect that makes every

data analysis a process. Others have named this data analysis

process the data analysis cycle (Hardin et al., 2015). This

includes the overall assessment of the problem, experimental

design, data acquisition, data cleaning, data transformation,

modeling and interpretation, prediction and the performance

of an exploratory as well as conﬁrmatory analysis. Especially

the exploratory data analysis part (Tukey, 1977) makes it clear

that there is an interaction process between the data analyst

and the data under investigation, which follows a systematic

approach, but is strongly inﬂuenced by the feedback of previous

analysis steps making a static description from the outset usually

impossible. Metaphorically, the result of a data analysis process

is like a cocktail having a taste that is beyond its constituting

ingredients. The process character of the analysis is also an

expression of the fact that, typically, there is not just one method

that allows answering a complex, domain speciﬁc question

but the consecutive application of multiple methods allows

achieving this. From this perspective it becomes also clear why

statistical inference is more than the mere understanding of

the technicalities of one method but that the output of one

method forms the input of another method which is requires the

sequential understanding of decision processes.

From the above description, a layman may think that statistics

is data science, because the above elements can also be found

in statistics. However, this is not true. It is more what statistics

could be! Interestingly, we think it is fair to say that some (if

Frontiers in Genetics | www.frontiersin.org 2February 2016 | Volume 7 | Article 12

Emmert-Streib et al. Data Science

not all) of the founders of statistics, e.g., Fisher or Pearson can

be considered as data scientists because they (1) analyzed and

showed genuine interest in a large number of diﬀerent data types

and their underlying phenomena, (2) possessed mathematical

skills to develop new methods for their analysis, and (3) gathered

large amounts of data and crunched numbers (by human labor).

From this perspective one may wonder how statistics, that started

out as data science, could end at a diﬀerent point? Obviously,

there had to be a driving force during the maturation of the

ﬁeld that prevented statistics from continuing along its original

principles. We don’t think this is due to the lack of creativity

or insight of statisticians that didn’t recognize this deviation,

instead, we think that the institutionalization of statistics or

science in general, e.g., by the formation of departments of

statistics and their bureaucratic management as well as an era

of slow progression in the development of data generation

technologies, e.g., comparing the periods 1950–1970 with 2000-

present, are major sources for this development. Given that,

naturally, the beginning of every scientiﬁc discipline is not only

indicated by the lack of a well deﬁned description of the ﬁeld

and its goals, but rather by a collection of people who share a

common vision. Therefore, it is clear that a formalization of a

ﬁeld leads inevitably to a restriction in its scope in order to form

clear boundaries to other disciplines. In addition, the increasing

introduction of managemental structures in universities and

departments is an accelerating factor that led to a further

reduction in ﬂexibility and tolerability of the variability in

individual research interests. The latter is certainly true for every

discipline, not just statistics.

More speciﬁc to statistics is the fact that the technological

progress between the 1930s and 1980s was rather slow compared

to the developments within the last 30 years. These periods of

stasis may have given the impression that developing ﬁxed sets

of methods is suﬃcient to deal with all possible problems. This

may also explain the wide spread usage of software packages like

SPSS or SAS among statisticians despite their limitations of not

being fully ﬂexible programming languages (Turing machines

Turing, 1936), oﬀering exactly these ﬁxed sets of methods.

In combination with the adaptation of the curriculum toward

teaching students the usage of such software packages instead

of programming languages, causes nowadays problems since

the world changed and every year appear new technologies

that are accompanied by data types with new and challenging

characteristics; not to mention the integration of such data sets.

Overall, all of these developments made statistics more applied

and also less ﬂexible, opening in this way the door for a new ﬁeld

to ﬁll the gap. This ﬁeld is data science.

A contribution that should not be underestimated in making,

e.g., computational biology a data science is the development

of the statistical programming language R pioneered by

Robert Gentleman and Ross Ihaka (Altschul et al., 2013)

and the establishment of the package repository Bioconductor

(Gentleman et al., 2004). In our opinion the key to success

is the ﬂexibility of R being particularly suited for a statistical

data analysis, yet having all features of a multi-purpose

language and its interface to integrate programs written in other

major languages like C++ or Fortran. Also, the license free

availability for all major operating systems, including Windows,

Apple and Linux, makes R an enabler for all kinds of data

related problems that is in our opinion currently without

rivalry.

Interestingly, despite the fact that computational biology has

currently all attributes that makes it a data science, the case of

statistics teaches us that this does not have to be this way forever.

A potential danger to the ﬁeld is to spend too much eﬀort on

the development of complex and complicated algorithms when in

fact a solution can be achieved by simple methods. Furthermore,

the iterative improvement of methods that lead only to marginal

improvements of results consumes large amounts of resources

that distract from the original problem buried within given data

sets. Last but not least, the increasing institutionalization of

computational biology at universities and research centers may

lead to a less ﬂexible ﬁeld as discussed above. All such inﬂuences

need to be battled because otherwise, in a couple of years from

now, people may wonder how computational biology could leave

the trajectory from being a data science.

AUTHOR CONTRIBUTIONS

FS conceived the study. FS, SM, and MD wrote the paper and

approved the ﬁnal version.

FUNDING

MD thanks the Austrian Science Funds for supporting this work

(project P26142). MD gratefully acknowledges ﬁnancial support

from the German Federal Ministry of Education and Research

(BMBF) (project RiKoV, Grant No. 13N12304).

ACKNOWLEDGMENTS

For professional proof reading of the manuscript we would like

to thank Bárbara Macías Solís.

REFERENCES

Altschul, S., Demchak, B., Durbin, R., Gentleman, R., Krzywinski, M., Li, H.,

et al. (2013). The anatomy of successful computational biology software. Nat.

Biotechnol. 31, 894–897. doi: 10.1038/nbt.2721

Conway, D. (2003). The data science venn diagram. Available online at: http://

www.drewconway.com/zia/2013/3/26/the-data- science-venn- diagram

Efron, B. (2003). “The statistical century,” in Stochastic Musings: Perspectives

From the Pioneers of the Late 20th Century, ed J. Panaretos (New York, NY:

Psychology Press), 29–44.

Gentleman, R. C., Carey, V. J., Bates, D. M., Bolstad, B., Dettling, M., Dudoit,

S., et al. (2004). Bioconductor: open software development for computational

biology and bioinformatics. Genome Biol. 5, R80. doi: 10.1186/gb-2004-5-

10-r80

Frontiers in Genetics | www.frontiersin.org 3February 2016 | Volume 7 | Article 12

Emmert-Streib et al. Data Science

Hardin, J., Hoerl, R., Horton, N. J., Nolan, D., Baumer, B., Hall-Holt, O., et al.

(2015). Data science in statistics curricula: preparing students to ‘think with

data’. Am. Stat. 69, 343–353. doi: 10.1080/00031305.2015.1077729

Press, G. (2013). A very short history of data science. Forbes. Available online

at: http://www.forbes.com/sites/gilpress/2013/05/28/a-very- short-history-of-

data-science/#2715e4857a0b405b5aa869fd

Quackenbush, J. (2011). The Human Genome: The Book of Essential Knowledge.

New York, NY: Imagine Publishing.

The Cancer Genome Atlas Research Network (2008). Comprehensive genomic

characterization deﬁnes human glioblastoma genes and core pathways. Nature

455, 1061–1068. doi: 10.1038/nature07385

Tukey, J. (1977). Exploratory Data Analysis. New York, NY: Addison-Wesley.

Turing, A. (1936). On computable numbers, with an application to the

entscheidungsproblem. Proc. Lond. Math. Soc. 2, 230–265.

Vapnik, V. N. (1995). The Nature of Statistical Learning Theory. New York, NY:

Springer.

Conﬂict of Interest Statement: The authors declare that the research was

conducted in the absence of any commercial or ﬁnancial relationships that could

be construed as a potential conﬂict of interest.

article distributed under the terms of the Creative Commons Attribution License

(CC BY). The use, distribution or reproduction in other forums is permitted,

provided the original author(s) or licensor are credited and that the original

publication in this journal is cited, in accordance with accepted academic practice.

No use, distribution or reproduction is permitted which does not comply with these

terms.

Frontiers in Genetics | www.frontiersin.org 4February 2016 | Volume 7 | Article 12

Data Science and Interdisciplinarity

Chapter

Jan 2022

In parallel with the progression of technology, the tourism industry has been continuously confronted with a large amount of data that needs to be systematically analyzed in order to gain significant insights into the science and business sectors. Data science has emerged as an interdisciplinary field where specific competencies from different sub-disciplines come together. This poses far-reaching challenges for both researchers and practitioners alike. To unlock the pillars of data science research and provide a guideline for relevant stakeholders in tourism, this chapter aims to conceptualize the core competencies needed in the data science process. More specifically, it will start with a discussion regarding the interplay between computer science, mathematics and statistics, and domain knowledge. Next, the procedure of data science will be classified into seven distinct phases: (1) topic formulation and relevance for academia and industry, (2) data access and data collection, (3) data pre-processing, (4) feature engineering, (5) analysis, (6) model evaluation and model tuning, and (7) interpretation of results. This chapter will review each stage in depth and evaluate the corresponding level of knowledge and competencies required for each phase. Finally, current implications and potential future directions of data science in the tourism industry will be discussed.

Ensuring the Robustness and Reliability of Data-Driven Knowledge Discovery Models in Production and Manufacturing

Article

Full-text available

Jun 2021

The Cross-Industry Standard Process for Data Mining (CRISP-DM) is a widely accepted framework in production and manufacturing. This data-driven knowledge discovery framework provides an orderly partition of the often complex data mining processes to ensure a practical implementation of data analytics and machine learning models. However, the practical application of robust industry-specific data-driven knowledge discovery models faces multiple data- and model development-related issues. These issues need to be carefully addressed by allowing a flexible, customized and industry-specific knowledge discovery framework. For this reason, extensions of CRISP-DM are needed. In this paper, we provide a detailed review of CRISP-DM and summarize extensions of this model into a novel framework we call Generalized Cross-Industry Standard Process for Data Science (GCRISP-DS). This framework is designed to allow dynamic interactions between different phases to adequately address data- and model-related issues for achieving robustness. Furthermore, it emphasizes also the need for a detailed business understanding and the interdependencies with the developed models and data quality for fulfilling higher business objectives. Overall, such a customizable GCRISP-DS framework provides an enhancement for model improvements and reusability by minimizing robustness-issues.

Evaluation of Regression Models: Model Assessment, Model Selection and Generalization Error

Article

Full-text available

Mar 2019

When performing a regression or classification analysis, one needs to specify a statistical model. This model should avoid the overfitting and underfitting of data, and achieve a low generalization error that characterizes its prediction performance. In order to identify such a model, one needs to decide which model to select from candidate model families based on performance evaluations. In this paper, we review the theoretical framework of model selection and model assessment, including error-complexity curves, the bias-variance tradeoff, and learning curves for evaluating statistical models. We discuss criterion-based, step-wise selection procedures and resampling methods for model selection, whereas cross-validation provides the most simple and generic means for computationally estimating all required entities. To make the theoretical concepts transparent, we present worked examples for linear regression models. However, our conceptual presentation is extensible to more general models, as well as classification problems.

Defining a Digital Twin: A Data Science-Based Unification

Article

Full-text available

Aug 2023

Frank Emmert-Streib

The concept of a digital twin (DT) has gained significant attention in academia and industry because of its perceived potential to address critical global challenges, such as climate change, healthcare, and economic crises. Originally introduced in manufacturing, many attempts have been made to present proper definitions of this concept. Unfortunately, there remains a great deal of confusion surrounding the underlying concept, with many scientists still uncertain about the distinction between a simulation, a mathematical model and a DT. The aim of this paper is to propose a formal definition of a digital twin. To achieve this goal, we utilize a data science framework that facilitates a functional representation of a DT and other components that can be combined together to form a larger entity we refer to as a digital twin system (DTS). In our framework, a DT is an open dynamical system with an updating mechanism, also referred to as complex adaptive system (CAS). Its primary function is to generate data via simulations, ideally, indistinguishable from its physical counterpart. On the other hand, a DTS provides techniques for analyzing data and decision-making based on the generated data. Interestingly, we find that a DTS shares similarities to the principles of general systems theory. This multi-faceted view of a DTS explains its versatility in adapting to a wide range of problems in various application domains such as engineering, manufacturing, urban planning, and personalized medicine.

A Roadmap for Building Data Science Capacity for Health Discovery and Innovation in Africa

Article

Full-text available

Oct 2021

Technological advances now make it possible to generate diverse, complex and varying sizes of data in a wide range of applications from business to engineering to medicine. In the health sciences, in particular, data are being produced at an unprecedented rate across the full spectrum of scientific inquiry spanning basic biology, clinical medicine, public health and health care systems. Leveraging these data can accelerate scientific advances, health discovery and innovations. However, data are just the raw material required to generate new knowledge, not knowledge on its own, as a pile of bricks would not be mistaken for a building. In order to solve complex scientific problems, appropriate methods, tools and technologies must be integrated with domain knowledge expertise to generate and analyze big data. This integrated interdisciplinary approach is what has become to be widely known as data science. Although the discipline of data science has been rapidly evolving over the past couple of decades in resource-rich countries, the situation is bleak in resource-limited settings such as most countries in Africa primarily due to lack of well-trained data scientists. In this paper, we highlight a roadmap for building capacity in health data science in Africa to help spur health discovery and innovation, and propose a sustainable potential solution consisting of three key activities: a graduate-level training, faculty development, and stakeholder engagement. We also outline potential challenges and mitigating strategies.

From the Digital Data Revolution toward a Digital Society: Pervasiveness of Artificial Intelligence

Article

Full-text available

Mar 2021

Frank Emmert-Streib

Technological progress has led to powerful computers and communication technologies that penetrate nowadays all areas of science, industry and our private lives. As a consequence, all these areas are generating digital traces of data amounting to big data resources. This opens unprecedented opportunities but also challenges toward the analysis, management, interpretation and responsible usage of such data. In this paper, we discuss these developments and the fields that have been particularly effected by the digital revolution. Our discussion is AI-centered showing domain-specific prospects but also intricacies for the method development in artificial intelligence. For instance, we discuss recent breakthroughs in deep learning algorithms and artificial intelligence as well as advances in text mining and natural language processing, e.g., word-embedding methods that enable the processing of large amounts of text data from diverse sources such as governmental reports, blog entries in social media or clinical health records of patients. Furthermore, we discuss the necessity of further improving general artificial intelligence approaches and for utilizing advanced learning paradigms. This leads to arguments for the establishment of statistical artificial intelligence. Finally, we provide an outlook on important aspects of future challenges that are of crucial importance for the development of all fields, including ethical AI and the influence of bias on AI systems. As potential end-point of this development, we define digital society as the asymptotic limiting state of digital economy that emerges from fully connected information and communication technologies enabling the pervasiveness of AI. Overall, our discussion provides a perspective on the elaborate relatedness of digital data and AI systems.

Human team behavior and predictability in the massively multiplayer online game WOT Blitz

Article

Aug 2023

Massively multiplayer online games (MMOGs) played on the Web provide a new form of social, computer-mediated interactions that allow the connection of millions of players worldwide. The rules governing team-based MMOGs are typically complex and non-deterministic giving rise to an intricate dynamical behavior. However, due to the novelty and complexity of MMOGs their behavior is understudied. In this paper, we investigate the MMOG World of Tanks (WOT) Blitz by using a combined approach based on data science and complex adaptive systems. We analyze data on the population level to get insight into organizational principles of the game and its game mechanics. For this reason, we study the scaling behavior and the predictability of system variables. As a result, we find a power-law behavior on the population level revealing long-range interactions between system variables. Furthermore, we identify and quantify the predictability of summary statistics of the game and its decomposition into explanatory variables. This reveals a heterogeneous progression through the tiers and identifies only a single system variable as key driver for the win rate.

Fen Kitap

Book

Full-text available

Feb 2022

In this study, the investigations of the non-Newtonian fluid and heat transfer is considered as an exponentially stretching surface in the occurrence of porous magnetic field due to its diversified applications in medical and engineering disciplines.The behavior of non-Newtonian fluid has been characterized as Casson fluid model.The major parameters like thermal radiation and viscous dissipation are incorporated in the energy equation. By considering governing equations of the mathematical model as a platform partial differential equations have been reduced to ordinary differential equations by using similarity transformations with proper boundary conditions.BVP4C software technique is used for finding the solutions and results are presented in the form of tables and graphs for the momentum and energy equations.After simulation, it is found that the heat transfer rate decreases with higher values of the magnetic field and radiation, and the temperature profile of the Casson fluid flow is directly related to the Casson fluid parameters. Finally, it is observed that parameters such as Casson fluid, Prandtl number, and Eckert number are stable at the point of convergence.

Passing the Data Baton: A Retrospective Analysis on Data Science Work and Workers

Conference Paper

Oct 2020

Passing the Data Baton: A Retrospective Analysis on Data Science Work and Workers

Article

Oct 2020

Data science is a rapidly growing discipline and organizations increasingly depend on data science work. Yet the ambiguity around data science, what it is, and who data scientists are can make it difficult for visualization researchers to identify impactful research trajectories. We have conducted a retrospective analysis of data science work and workers as described within the data visualization, human computer interaction, and data science literature. From this analysis we synthesis a comprehensive model that describes data science work and breakdown to data scientists into nine distinct roles. We summarise and reflect on the role that visualization has throughout data science work and the varied needs of data scientists themselves for tooling support. Our findings are intended to arm visualization researchers with a more concrete framing of data science with the hope that it will help them surface innovative opportunities for impacting data science work.

Data Science in Statistics Curricula: Preparing Students to "Think with Data"

Article

Full-text available

Oct 2014

A growing number of students are completing undergraduate degrees in statistics and entering the workforce as data analysts. In these positions, they are expected to understand how to utilize databases and other data warehouses, scrape data from Internet sources, program solutions to complex problems in multiple languages, and think algorithmically as well as statistically. These data science topics have not traditionally been a major component of undergraduate programs in statistics. Consequently, a curricular shift is needed to address additional learning outcomes. The goal of this paper is to motivate the importance of data science proficiency and to provide examples and resources for instructors to implement data science in their own statistics curricula. We provide case studies from seven institutions. These varied approaches to teaching data science demonstrate curricular innovations to address new needs. Also included here are examples of assignments designed for courses that foster engaging of undergraduates with data and data science.

Erratum: The anatomy of successful computational biology software

Article

Full-text available

Jun 2014

Creators of software widely used in computational biology discuss the factors that contributed to their success

Comprehensive genomic characterization defines human glioblastoma genes and core pathways

Article

Full-text available

Oct 2008
NATURE

Human cancer cells typically harbour multiple chromosomal aberrations, nucleotide substitutions and epigenetic modifications that drive malignant transformation. The Cancer Genome Atlas (TCGA) pilot project aims to assess the value of large-scale multi-dimensional analysis of these molecular characteristics in human cancer and to provide the data rapidly to the research community. Here we report the interim integrative analysis of DNA copy number, gene expression and DNA methylation aberrations in 206 glioblastomas-the most common type of adult brain cancer-and nucleotide sequence aberrations in 91 of the 206 glioblastomas. This analysis provides new insights into the roles of ERBB2, NF1 and TP53, uncovers frequent mutations of the phosphatidylinositol-3-OH kinase regulatory subunit gene PIK3R1, and provides a network view of the pathways altered in the development of glioblastoma. Furthermore, integration of mutation, DNA methylation and clinical treatment data reveals a link between MGMT promoter methylation and a hypermutator phenotype consequent to mismatch repair deficiency in treated glioblastomas, an observation with potential clinical implications. Together, these findings establish the feasibility and power of TCGA, demonstrating that it can rapidly expand knowledge of the molecular basis of cancer.

Bioconductor: Open Software Development for Computational Biology and Bioinformatics

Article

Full-text available

Sep 2004

The Bioconductor project is an initiative for the collaborative creation of extensible software for computational biology and bioinformatics. The goals of the project include: fostering collaborative development and widespread use of innovative software, reducing barriers to entry into interdisciplinary scientific research, and promoting the achievement of remote reproducibility of research results. We describe details of our aims and methods, identify current challenges, compare Bioconductor to other open bioinformatics projects, and provide working examples.

On computable numbers, with an application to the Entscheidungsproblem

Article

Jan 1936

A.M. Turing

The Nature of Statistical Learning Theory

Chapter

Jan 2000

Vladimir N. Vapnik

In the history of research of the learning problem one can extract four periods that can be characterized by four bright events: (i) Constructing the first learning machines, (ii) constructing the fundamentals of the theory, (iii) constructing neural networks, (iv) constructing the alternatives to neural networks.

Exploratory Data Analysis

Article

Sep 1979

The Nature Of Statistical Learning Theory

Book

Jan 1995

Vladimir N. Vapnik

Setting of the learning problem consistency of learning processes bounds on the rate of convergence of learning processes controlling the generalization ability of learning processes constructing learning algorithms what is important in learning theory?.

Erratum: Comprehensive genomic characterization defines human glioblastoma genes and core pathways (Nature (2008) 455 (1061-1068) DOI: 10.1038/nature07385)

Article