Bachelor - Data science - IT Universitetet i København

The programme comprises mandatory study activities worth 135 ECTS points, track specific study activities worth 15 ECTS, optional study activities worth 15 ECTS points, and a bachelor project worth 15 ECTS points.
The courses listed below constitute the core elements of the programme and are worth 150 ECTS points.

- Introduction to Data Science and Programming (15 ECTS)
- Applied Statistics (7,5 ECTS)
- Data Science in Research, Business and Society (7,5 ECTS)
- First Year Project (15 ECTS)
- Algorithms and Data Structures (7,5 ECTS)
- Linear Algebra and Optimization (7,5 ECTS)
- Machine Learning (15 ECTS)
- Data Management (7,5 ECTS)
- Network Analysis (7,5 ECTS)
- Second Year Project (15 ECTS)
- Data Visualization and Data-driven Decision Making (7,5 ECTS)
- Technical Communication (7,5 ECTS)
- Security and Privacy (7.5 ECTS)
- Reflections on Data Science (Videnskabsteori)(7,5 ECTS points)
- Bachelor Project (15 ECTS points)

There will additionally be two elective courses, and mandatory courses depending on the student’s “track”.
For a graphical overview of the programme please see Course plan in upload: Dokumentation af efterspørgsel på uddannelsesprofil.

Assisted collaboration with industry.  Projects that connect to real data and questions are an integral part of the programme, developed in collaboration with external partners. The design of the projects was discussed with employers at the employers’ meeting 30 May at ITU and all the employers present expressed great interest in participating and contributing in collaborative projects (Minutes Employers Meeting. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil).

Qualification for ITU’s MSc programmes.  There are two specialisations so called "tracks" with slightly different mandatory courses: Technical and Business. The Technical track gives right to admission to the MSc in Software Development and the Business track gives right to admission to the MSc in Digital Innovation & Management.
Technical track: Programming and Software Engineering (7,5 ECTS) and Large-scale Data Analysis (7,5 ECTS).
Business track: Organisation Theory (7,5 ECTS) and Process Management (7,5 ECTS).

Exchange opportunities. The 5th semester is well suited for a stay abroad. We expect that it will be possible to find courses matching the two mandatory courses and one track course in most universities.

Study environment.  The programme is accompanied by academic support activities such as the study café and mentoring programmes that promote a culture of engagement and a deep approach to learning among students.

Short course descriptions:
- Introduction to Data Science and Programming. An introduction to Data Science, including basic data processing, visualisation, data sampling, prediction, inference, and probability, coupled with a hands-on introductory programming class.  After the course, students should be able to do basic data processing and analysis using Python as a tool.  Possible books: (1a) Computational and inferential thinking [1] or (1b) Data Science from Scratch [2], together with (2)  Introduction to Programming in Python [3].
- Applied Statistics. An introduction to statistical learning, including basic probability theory, statistical modeling and inference. Examples are implemented in the functional statistical programming language R. After the course students are able to make valid statistical analyses, and identify statistical flaws in data analysis. Possible book: An Introduction to Statistical Learning with Applications in R [4].
- Data Science in Research, Business, and Society. A substantive discussion of applications and opportunities for Data Science with examples of real world problems coupled with an overview of field-specific research approaches for addressing these problems. After the course students are able to develop empirical research questions and define empirical research problems, discuss organizational processes and societal concerns relevant to Data Science. Possible books: Big data, little data, no data: scholarship in the networked world [5] and The data revolution: Big data, open data, data infrastructures and their consequences [6].
- First Year Project. Data Science project with data sets and objectives proposed by (in-house) domain experts. Composed, for example, of a series of seven 2-week mini-projects. A coordinator will run the course managing domain experts in business, games, digital humanities, and IoT. After the project students are able to translate diverse problem settings into a well-defined data analysis problem, implement a solution to the problem, and translate the findings back to the problem domain.
- Algorithms and Data Structures. The course introduces structures that allow efficient organization and retrieval of data, frequently used algorithms, and generic techniques for modeling, understanding, and solving algorithmic problems. After the course students are able to make informed choices of algorithms and data structures for data analysis based on knowledge of their resource requirements. Possible book: Algorithms and data structures: The Basic Toolbox [7].
- Linear Algebra and Optimization. The goal of this course is to provide firm foundations in linear algebra and optimization techniques. After the course students are able to analyze and solve optimization problems arising in areas such as machine learning. Possible book: Fundamentals of Linear Algebra and Optimization [8].
- Machine Learning. Comprehensive overview of machine learning methods and their application in practice. After the course students are able to choose appropriate machine learning methods for a given setting, and make simple modifications to machine learning methods to improve effectiveness. Possible book: Pattern Recognition and Machine Learning [9].
- Data Management. Introduction to fundamental concepts of data modelling, storage, organization, querying and transactions. After the course students are able to choose a data model and a database system that fits a given problem definition, and implement analytics solutions (e.g. in SQL) that pull data from a database system. Possible book: Patterns in Data Management [10].
- Second Year Project. Data Science project with topic, data sets and domain expertise proposed by external stakeholders (e.g., analysis of patterns of traffic around schools in the city of Copenhagen, in partnership with Københavns Kommune). After the project students are able to combine various tools and methods to get new insights in a complex domain. This can also be a chance to get experience with widely used data analysis tools such as SAS.
- Reflections on Data Science. Epistemology and ethics of quantitative reasoning. After the course students will be able to analyze a problem of interest that touches upon the relationships between Data Science and its context (may it be of political, ethical, philosophical, historical or societal nature). Reading material: a compendium of components of different books/articles to cover content relevant specifically to Data Science.
- Technical Communication. Skills for communicating analytics findings to non-technical stakeholders. After the course students will be able to choose and use effective means for communicating, orally and in writing, technical findings at all levels of understanding. Possible book: Technical Communication [11].
- Large-scale Data Analysis (Technical track). Overview of the models, systems and algorithms for analyzing large volumes of data, both in batch and in real-time. After the course students are able to use frameworks for large-scale data analysis and reason about the correctness and efficiency of their implementations. Possible books: (1) Map-Reduce Design Patterns [12], and (2) Big Data Analytics Beyond Hadoop: Real-Time Applications with Storm, Spark and More Hadoop Alternatives [13].
- Network Analysis. Basics of graph theory, network analysis methodology and social network analytics approaches. After the course students are able to describe, measure and analyze social networks, extract and visualize network data and apply this knowledge to real-world problems. Possible book: Introduction to Social Network Analysis with R [14], supplemented by Networks, Crowds, and Markets [15].
- Data Visualization and Data-driven Decision Making. Basic introduction to data visualisation techniques for exploratory, analytical and presentation purposes. Introduction to data-driven decision making strategies, demands and concerns in public and private organizations. After the course students are able to define necessary resources for a particular decision-making task, identify the avenues for obtaining data and insights and discuss potential legal and societal concerns with this process, to select appropriate data visualisation techniques and to produce visualisation of relevant data for stakeholders. Possible books: (1) Visual Display of Quantitative Information [16], and (2) Data Science for Business [17].
- Software Engineering and Programming (Technical track). Introduction to the Java language and modular software development aimed at students that already know a programming language (such as Python).  After the course, students are fluent in Java and modular software development.
- Security and Privacy. Techniques for information security and privacy in a networked world. The course will include a basic overview of the General Data Protection Regulation as well as an overview of legal, technical and societal concerns around data collection, data processing and data-driven decision making. After the course students are able to design solutions that comply with regulations, based on security and privacy best practices. Possible books: (1) Elementary Information Security [18] and (2) Privacy in context: Technology, policy, and the integrity of social life [19].
- Organization Theory (Business track). Introduction to the field of organization theory, addressing topics such as work, action, practice and complexity and how to deal with such topics analytically. After the course, students will be able to demonstrate how information technologies and data affect organizational activities and practices. Possible reading material: Images of Organization [20] or Organization theory [21], together with a selection of journal papers.
- Process Management (Business track). Introduction to the analysis, design, implementation, and improvement of organizational work processes and supporting IT systems. After the course, students will master the phases of the process management lifecycle, from process identification over modeling, analysis, redesign, automation, to process monitoring and performance management.  Students will have engaged with real-world case studies related to process-oriented management methodologies like Six Sigma and Lean Management. Possible book: Fundamentals of BPM [22].

Examples of electives:
- Deep learning
- Natural language processing
- Multimedia analytics
- Business foundations
- Algorithm design
- Network society
- Digital material and social media
- Functional programming

Course book references
[1] Adhikari, Ani and John DeNero. Computational and Inferential Thinking. https://www.gitbook.com/book/ds8/textbook/details
[2] Grus, Joel. (2015).  Data Science from Scratch.  First Principles with Python. O’Reilly. http://shop.oreilly.com/product/0636920033400.do
[3] Sedgewick, Robert, Kevin Wayne and Robert Dondero. (2015).  Introduction to Programming in Python. Pearson. http://introcs.cs.princeton.edu/python/home/
[4] James, Gareth, Daniela Witten, Trevor Hastie, and Robert Tibshirani. (2013). An Introduction to Statistical Learning with Applications in R.  Springer. http://www-bcf.usc.edu/~gareth/ISL/
[5] Borgman, C. L. (2015). Big data, little data, no data: scholarship in the networked world. MIT Press. https://mitpress.mit.edu/big-data-little-data-no-data
[6] Kitchin, R. (2014). The data revolution: Big data, open data, data infrastructures and their consequences. Sage. https://thedatarevolutionbook.wordpress.com/
[7] Melhorn, Kurt and Peter Sanders. (2008).  Algorithms and Data Structures.  Springer.  http://www.springer.com/gp/book/9783540779773
[8] Gallier, Jean. (2016).  Fundamentals of Linear Algebra and Optimization.  http://www.seas.upenn.edu/~cis515/linalg.pdf
[9] Bishop, Christopher.  (2006).  Pattern Recognition and Machine Learning.  Springer. http://www.springer.com/gp/book/9780387310732
[10] Dittrich, Jens. (2015). Patterns in Data Management. https://infosys.uni-saarland.de/datenbankenlernen/Patterns_In_Data_Management_Preview.pdf
[11] Markel, Mike. (2012).  Technical Communication. 10th ed. Bedford/St. Martin's.
[12] Shook, Adam and Donald Miner. (2012). MapReduce Design Patterns.  O’Reilly. https://www.safaribooksonline.com/library/view/mapreduce-design-patterns/9781449341954/
[13] Srinivas Agneeswaran, Vijay. (2014). Big Data Analytics Beyond Hadoop:  Real-Time Applications with Storm, Spark, and More Hadoop Alternatives.  Pearson FT. https://www.safaribooksonline.com/library/view/big-data-analytics/9780133838268/
[14] Bojanowski, Michal. (2017).  Introduction to Social Network Analysis with R.  Wiley. http://eu.wiley.com/WileyCDA/WileyTitle/productCd-1118456041.html
[15] Easley, David and Jon Kleinberg. (2010). Networks, Crowds, and Markets: Reasoning About a Highly Connected World.  Cambridge University Press. https://www.cs.cornell.edu/home/kleinber/networks-book/
[16] Tufte, Edward R. The Visual Display of Quantitative Information.  https://www.edwardtufte.com/tufte/books_vdqi
[17] Provost, Foster and Fawcett, Tom. (2013).  Data Science for Business: What You Need to Know about Data Mining and Data-Analytic Thinking.  O’Reilly Media. http://shop.oreilly.com/product/0636920028918.do
[18] Smith, Richard E. (2016). Elementary Information Security. 2nd ed. Jones & Bartlett. http://www.jblearning.com/catalog/9781284055931/
[19] Nissenbaum, Helen. (2009).  Privacy in Context: Technology, Policy, and the Integrity of Social Life.  Stanford University Press.http://www.sup.org/books/title/?id=8862
[20] Morgan, Gareth. (2007). Images of Organization.  Sage. https://uk.sagepub.com/en-gb/eur/images-of-organization/book229704
[21] Hatch, Mary Jo and Cunliffe, Ann L. (2013). Organization Theory. Oxford University Press. http://www.oupcanada.com/catalog/9780199640379.html
[22] Dumas, Maron, Marcello La Rosa, Jan Mendling, and Hajo A. Reijers. (2013). Fundamentals of Business Process Management. Springer. http://shop.oreilly.com/product/0636920033400.do

There is widespread and growing demand for data scientists on the Danish job market, as well as worldwide (particularly in USA and Europe).  Currently the demand for data scientists is so high that companies must seek candidates abroad (particularly, in the USA and the UK).  Federal reports in Denmark warn of this current and expected growth, with the lack of expertise in Data Science (also sometimes referred to as Big Data) consistently named as a central area for concern.  And it is an explicit recommendation from Danmarks Vækstråd to educate more data scientists (Danmarks Vækstråd, Anbefalinger vedr. datadreven udvikling og vækst. 2015).

Denmark is behind in the data game (emphasis is ours):
- “[...] Danmark er det mest digitale samfund i EU, og at virksomhedernes digitalisering på mange områder er helt i front. Fremadrettet er der dog også en række udfordringer, som skal tages alvorligt.  [...] Samtidig er virksomhederne bagud på anvendelsen inden for nyere digitale vækstområder – som fx dataanalyse”  (Erhvervs og Vækstministeriet: Redegørelse om Danmarks digitale vækst 2016 (a))

- “Sammenlignet med de bedste i Europa, herunder også øvrige nordeuropæiske lande, indsamler og analyserer danske virksomheder i mindre grad kundedata. I en digital tidsalder er data et afgørende råstof, der skaber nye forretningsmuligheder for erhvervslivet, og som virksomheder kan bruge til at optimere deres forretningsgang” (Erhvervs og Vækstministeriet: Redegørelse om Danmarks digitale vækst 2016 (b), p11)

- “[D]er inden for både energi, fødevarer og handel er store potentialer inden for big data. En mindre gruppe af frontløbervirksomheder er allerede begyndt at høste gevinsterne. Men generelt er potentialerne langt større end resultaterne i alle tre sektorer” (IRIS Group, Datadreven vækst i Danmark. 2015,  p3)

One source for this lagging behind is the lack of Data Science competencies in Denmark, as well as educational resources to produce candidates with these (emphasis is ours):

- "Der kan også [...] konstateres en række fælles træk i de barrierer, der gør det vanskeligt for virksomhederne at høste potentialerne. Det drejer sig bl.a. om usikkerhed, manglende indsigt i anvendelsesmuligheder, kompetencemangler, uensartede standarder, adgang til offentlige data og tilbageholdenhed ved at dele data med andre i værdikæden.
[...] udbredelsen af big data og datadreven forretningsudvikling kan stimuleres gennem målrettede forbedringer i en række offentlige vækstbetingelser. Det gælder på områder som forskning, uddannelse, regulering, prissætning på offentlige data og adgang til uvildig sparring.” (IRIS Group, Datadreven vækst i Danmark. 2015, p3)

- “The demand for data specialist skills will exceed the current supply of the labour market and the current capacity of education and training systems, requiring rapid adjustments in curricula and the skill sets of teachers and on-the-job workers.” (OECD, An OECD horizon scan of megatrends and technology trends in the context of future research policy. 2016, p53)

- “Det er bekymrende, at vi er bagud med at bruge digitale værktøjer til at producere smartere eller til at analysere data – som der er et meget stort forretningspotentiale i – bl.a. fordi vi mangler de rigtige medarbejdere.” --Adam Lebech, Director, CEO, DI Digital (DI: Stort behov for øget  dansk digitalisering. 2016)

- “Lige nu har vi en masse dygtige folk, der sidder og arbejder med det data, vi indhenter. Men det er folk, som er uddannet i noget andet, og det vil sige, at vi ikke har nogen uddannet i den faglighed, som en data scientist vil have – og den faglighed vil vi gerne have.” -- (Lisbeth Nielsen, Direktør, Sundhedsdatastyrelsen. 2016)

ITU’s BSc in Data Science programme would provide an important and much needed contribution to Danish Data Science competences.  The contribution will be unique: the first Bachelor of its kind, strong on Data Science related technical and mathematical skills, but with a central real-world applications focus.  ITU stands uniquely capable to make this contribution to the comprehensive development of Data Science competencies in Denmark, due to its de facto interdisciplinary nature, especially on the frontier between technology and society.

Outline of ITU’s research into the need for a BSc in Data Science.
ITU’s research into the need for the new BSc in Data Science has been both qualitative and quantitative.
1. Qualitative research.
- The Rambøll report.  ITU enlisted Rambøll Management Consulting to conduct a thorough independent qualitative investigation into the education demand within Denmark both in terms of the following:
a) The current and foreseeable required skills and competencies of data scientists as described by industry leaders in Denmark. (Cf. Section Hvilke aftagere/aftagerorganisationer har været inddraget i behovsundersøgelsen for a list of responants.)
b) The education landscape in Denmark with respect to these required skills and competencies, and how well this landscape meets the needs of the demand.
- Employers panel.  ITU has assembled its own panel of industry leaders in Denmark to help form the content of the BSc in Data Science programme.  Documentation from these meetings takes the form of minutes.

2. Quantitative research.
- Job listing analysis. ITU conducted a quantitative analysis of the Danish job market in terms of actual existing jobs for which Bachelors would be qualified.
All corresponding reports are attached as supplementary documentation. In addition, all data extracted and analysed for the quantitative report is available at https://owncloud.itu.dk/index.php/s/7ju4GeEnw2k5zPN . We now provide a short summary of our findings.

❖ The Rambøll Report (May 2016). In order to confirm that the global trend of increasing demand for data scientists is also manifest in Denmark, in February 2016, ITU enlisted Rambøll Management Consulting (Rambøll) to carry out an independent comprehensive qualitative investigation - including intense study of related current education programmes in Denmark, official federal reports, and direct employer feedback from a selection of leaders from the most influential companies within Denmark - into:
1. the current educational offerings within Data Science in Denmark (Section 3: “Delopgave I”), and
2. the expected size of the Danish labour within Data Science, and in particular with a 5-10 year horizon (Section 4: “Delopgave II”).
The resulting report, Markedsundersøgelse for data science is uploaded with this application. We now present a summary of the relevant main conclusions, and explain how ITU’s proposed Bachelor’s programme relates (Cf. Rambøll-Markedsundersøgelse. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil, p21).

1. Few Danish educations comprise the right combination of competencies for becoming a data scientist.  Rambøll conducted an exhaustive study of the educations in Denmark, finding that one Bachelor programme at DTU (Mathematics and Technology) could provide Bachelors with a sufficient cross-disciplinary understanding to serve as a basis for Data Science.  Further, Bachelor students of, for example, Technology, could tailor their programmes so that they could carry out Data Science.  However, there are no existing comprehensive educations leading to becoming a full-fledged data scientist in Denmark.
➤ ITU’s BSc in Data Science would be the first degree of its kind in Denmark, and we have been assured that Bachelors of this programme would be hireable directly out of the BSc programme both by leaders in Danish industry at our Employers Panel Meeting (Minutes Employers Meeting. 2016. Upload in Dokumentation af efterspørgsel på uddannelsesprofil) as well as via our own quantitative job listings analysis (Job Listings Analysis. 2016. Upload in Dokumentation af efterspørgsel på uddannelsesprofil). (See also Example Job Profiles, p. 14 in upload: Dokumentation af efterspørgsel på uddannelsesprofil.)  Furthermore, the competencies that the programme will develop are aligned those laid out by the Rambøll report (Cf. Course Plan, p2 in upload: Dokumentation af efterspørgsel på uddannelsesprofil).

2. Currently Data Science is a specialisation at the Master’s level.  The most relevant Danish educations within Data Science typically consist of a Bachelor education providing students with general Mathematics or Technology competencies, after which they individually can choose to specialise in Data Science at the Master’s level.
➤ Bachelor level studies focussing on Data Science will ensure a deeper, more versatile, and more comprehensive data scientist profile, which could not achieved by just a specialisation at the MSc level.  Moreover, ITU’s data scientists will have a strong analytical, results-oriented profile, with a view to solving real-world problems rather studying pure science for science.  The job prospectives for the resulting graduates correspond to Big Data versions of “Small Data” Market Analysts, Business Analysts, Games Analysts, or other Quantitative Social Scientists, where real-world questions are prioritised, and techniques to solve them are a means rather than an ends.  Still, the content of the programme will be high on sophisticated mathematical and technological skills required for Data Science, and it will be possible to complete a more technology-focussed Data Science Master’s (on the Bachelor’s Technical track, together with the MSc in Software Development).

3. The demand for data scientists is large and increasing.  This signal is clear, both nationally and internationally.  Typically data scientists have been employed within the Technology sector, but there is increasing demand for Data Science with in the Finance, Marketing, and Public sectors.  The demand is so great that many companies are forced to hire abroad.  All industry leaders surveyed agreed that this demand will continue in the 5-10 year spectrum, and that educating future data scientists cannot go fast enough.

❖ Employers Meeting Minutes.  Besides the feedback from relevant employers in the Rambøll Market investigation report ITU has also received direct feedback from other relevant employers at a half day meeting May 30, 2016 at the university. The employers present were industry and public sector leaders in Denmark. (Cf. Hvilke aftagere/aftagerorganisationer har været inddraget i behovsundersøgelsen.)  See upload: Dokumentation af efterspørgsel på uddannelsesprofil for the full meeting minutes (Minutes Employers Meeting. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil).
Before the meeting the employers received a 9-page programme proposal which included a description of the programme, graphic course plan, short course descriptions, examples of learning outcomes, possible MSc combinations together with examples of job profiles in relation to existing job positions, as well as the market investigation report Markedsundersøgelse for data science made by Rambøll (Rambøll-Markedsundersøgelse. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil).  At the meeting the demand for data scientists was discussed based on the demand in the industry and public sector represented by the employers present at the meeting. For more description of employers, see Section Hvilke aftagere/aftagerorganisationer har været inddraget i behovsundersøgelsen? The employers also provided feedback on the the content and structure of the programme in relation to the demand, which helped produced the programme contents and structure in its current form. For examples see Section Hvordan er det sikret, at den nye uddannelse matcher det påviste behov?

The employers stressed that they were positive towards the programme design, found it important that the programme is developed, and would like to hire both Bachelors and Graduates.
➤ Our Bachelors will have received a comprehensive foundation for practising Data Science, with focussed Data Science related scientific and technical skills and extensive work in their application.  The programme is an ideal fit for students whose main interest is data driven solutions and analysis of real-world problems, rather than the isolated technical skills.  We treat Data Science as a discipline in its own right, rather than as a specialisation in order to provide a complete Data Science education for versatile generalist data scientists that are capable of applying to a wide variety of real-world problems.  In addition, because of the unique profile of our intended student base for IT, the programme should be successful at attracting new young students to IT who would not have originally sought out a more technical or scientific education.

❖ Job listings analysis. ITU carried out quantitative studies regarding the current demand for data scientists in Denmark. The resulting technical report is provided (Job Listings Analysis. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil). The report includes two studies on Danish Data Science job listings data collected over 3-6 month periods in 2016:
1. a purely data-driven quantitative analysis Data Science related jobs advertisements available over 6 months of 2016, and
2. a hand-annotated quantitative analysis on Data Science related jobs advertisements for which graduates from the proposed program would be immediately qualified, over a period of 3 months in 2016.  The description of the 3 month study is given in Section Underbygget skøn over det samlede behov for dimittender as an estimate for the current demand in Denmark for data scientists.
The studies also provide concrete numbers the public and industry sectors in which the data scientist demand is manifested.  The main contribution of the studies is to show how the current demand for data scientists well exceeds the prospective number of graduates from the ITU’s proposed data scientist BSc programme, by several magnitudes, and that the BSc-MSc progressions proposed by ITU here are well-founded: there are a large majority of job advertisements related to business, and a solid presence of more technical Data Science jobs related to software.

How many graduates in Data Science should ITU produce?  There is ongoing and increasing demand for competent data scientists (Cf. Section Kort redegørelse for behovet for den nye uddannelse).  According to our the panel of Employers, educating data scientists can only go too slow (Minutes Employers Meeting. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil).

ITU carried out a quantitative study to gain an idea of the current demand for data scientists in Denmark.  Job advertisements from a central Danish job site were collected over a three month period (March, April, May of 2016), using a variety of keywords to ensure high recall on Data Science related jobs.  The advertisements were then hand-filtered for relevance to graduates from the proposed programme and annotated as (1) business-related, (2) software-related, or (3) other.  In total  270 were found to be job advertisements where a BSc in Data Science would be relevant (but a BSc in, say, software development would not be a good enough background).  About 65% of these are in business intelligence and 27% are software related.

The Rambøll report also provides some indicative figures for the future to this end.

- “[...] ITEK6  under Dansk Industri [opstiller] en prognose, som viser et udækket behov i Danmark  i 2020 på ca. 3.000 kandidater generelt i ITEK-branchen. ITEK konkluderer, at der skal tillægges en generel vækst inden for teknologibranchen, bl.a. i form af fremdrift i Big Data, hvilket skønnes at udgøre op til yderligere 3.000 kandidater (ITEK, 2015). I den tidligere nævnte undersøgelse af IDC tilkendegiver 13 pct. danske virksomheder, at de har planer om at ansætte en specialist inden for Big Data, mens 21 pct. udtrykker et behov for kompetencerne, men mangel på ressourcer (IDC, 2011). Ligeledes peger regeringens redegørelse for den digitale vækstplan på, at på trods af en stigning i antallet af it-uddannede, er det fortsat en udfordring at få dækket virksomhedernes behov inden for it, især i forhold til adgangen til specialiserede it-kompetencer, som fx it-udviklere og dataanalytikere (Regeringen, 2015).” (Rambøll-Markedsundersøgelse. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil, pp35-36).

Expected intake into the BSc in Data Science.  The ITU will start by admitting 50 students on the BSc in Data Science in 2017 and thereafter expects to increase the number of students according to the demand. Based on input from employers, Bachelors in Data Science may be able to get a job immediately following their BSc (Minutes Employers Meeting. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil). They may also choose to remain and combine their BSc with a relevant MSc, specialising for a particular field of application at the graduate level. (See the combination of MSc and job profiles in Section Sammenhæng med eksisterende uddannelser and in Example Job Profiles in upload: Dokumentation af efterspørgsel på uddannelsesprofil.)

The employers who participated in in the market investigation both at ITU and in the Rambøll report were carefully selected. The selection process took place in February and March 2016. ITU listed 60 potential employers including information on their affiliation, linkedin profile, area, and educational background and the list was used as basis for the selection for both the Rambøll report and the employers’ meeting at ITU.
In order to have the employer’s view represented in the selection, a Chairman for the coming group of employers was appointed. Kaare Brandt Petersen, Nordic Director for for Education & Academic from SAS Institute, assisted ITU in selecting a sub-group of employers to represent a broad range of employers for Data Science related positions in relevant industries and organisations.  The selection further ensured that both small and large companies were represented as well as both private and public sectors. Furthermore the educational background of the employers were also taken into account in order to ensure a broad perspective on Data Science.
The employers were approached in April 2016 with an initial description of the programme and invited to attend a meeting in May. The employers approached were very positive and seven out of the eight invited attended the meeting on May 30, 2016.
Since the May meeting employers are receiving updates on the continued development of the programme. At the end of August, ITU sent an update on which adjustments have been made based on the May meeting (see Section Hvordan er det sikret, at den nye uddannelse matcher det påviste behov? for a more detailed description). After the submission of the application for pre-qualification ITU will continue the dialogue with the employers and invite them to be part of a coming employers panel for the programme.

For the Markedundersøgelse for data science report ITU provided the above mentioned list of relevant employers to Rambøll and was in dialogue with Rambøll for the selection. In the appendices in the report it is stated: ...Da data science er et område, der er relevante indenfor flere og flere sektorer, har det været en prioritet at afdække et bredt felt potentielle aftagere indenfor forskellige brancher og sektorer… Respondenterne har primært haft stilling som chef eller direktør i virksomhedens teknologiske og analytiske afdelinger, mens respondenterne i de resterende interview ligeledes har været chefer, partnere eller direktører samt haft kendskab til data science på anden måde. (Rambøll-Markedsundersøgelse. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil p.57).

Please see Hvilke aftagere/aftagerorganisationer har været inddraget i behovsundersøgelsen in upload: Dokumentation af efterspørgsel på uddannelsesprofil for a full list of employers and their background.

In order to determine the relevant contents of the proposed BSc in Data Science programme, ITU carried out the research summarised in Section Kort redegørelse for behovet for den nye uddannelse and placed in upload: Dokumentation af efterspørgsel på uddannelsesprofil (i.e., the market investigation report made by Rambøll (Rambøll-Markedsundersøgelse. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil), an employers meeting with ongoing feedback (Minutes Employers Meeting. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil), and a data scientist job listings analysis (Job Listings Analysis. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil). The Rambøll report provided verification that the developed Data Science programme corresponded to industry needs in Denmark.  For active feedback on the alignment of the proposed Data Science programme with the needs of the employers, ITU gathered an employers panel for a half day meeting on May 30, 2016 at the university.  The employers present were industry and public sector leaders in Denmark. (See Section Hvilke aftagere/aftagerorganisationer har været inddraget i behovsundersøgelsen.)

Before the meeting the employers received a programme proposal and the Rambøll report.  At the meeting the demand for data scientists was discussed based on the demand in the industry and public sector represented by the employers present at the meeting. The employers also provided feedback on the the content and structure of the programme in relation to the demand, which helped produced the programme contents and structure in its current form.

The employers stressed that they were positive towards the programme design, found it important that the programme is developed, and would like to hire the graduates.  The employers from the meeting are continuously updated on the progress of the application and the structure of the programme, and are encouraged to provide as much feedback as possible to ensure success.

The feedback for suggested revisions from the employers was taken into account, and the programme contents and structure optimised for this.  We now give the optimisations made based on some of the issues raised by employers at the meeting:
1. The employers suggested we create a programme with two branches: one for data scientists that wish to achieve even stronger technical skills, and the other for those that focus on preparing to engage with businesses and organisations by understanding underlying business processes.  → The BSc in Data Science now consists of these two corresponding tracks: Technical and Business.
2. The employers expressed concern over students being exposed to real-world problems and learning the possible pitfalls or real data.  → The Data Science programme consists of several projects where students are meant to collaborate with industry.  This collaboration will be facilitated by ITU; ITU has successfully facilitated such collaboration in its other degree programmes for many years.  Moreover, the employers expressed interested in taking part in this collaboration.  The programme structure (See Course Plan in upload: Dokumentation af efterspørgsel på uddannelsesprofil p2) now includes icons to indicate the courses where students will be challenged by real-world problems and associated data as part of the curriculum.
3. The employers expressed concern that the mathematical and statistical content could be at least 1/3 of the programme.  → The programme does indeed contain at least 1/3 mathematics and statistics content.  This also more than delivers on the required mathematics competencies laid out by the Rambøll report (Rambøll-Markedsundersøgelse. 2016. In upload: Dokumentation af efterspørgsel på uddannelsesprofil, p22.  The programme structure (Course Plan p2, in upload: Dokumentation af efterspørgsel på uddannelsesprofil) now includes icons to indicate the courses heavy on mathematics/statistics content, to make this more explicit.  There are also icons to indicate heavy programming content.  Bachelors will become highly skilled in Data Science-related mathematics, statistics and programming.

The BSc in Data Science will be the first of its kind in Denmark and will set a standard for comprehensive Data Science education, as pointed out in Section Kort redegørelse for behovet for den nye uddannelse.  As laid out in Section Beskrivelse af uddannelsen, the study programme aims to educate versatile, strongly competent data scientists, ready for the job market.  Bachelors will have the opportunity to further specialise for a particular field of application at the graduate level.  Hence the full study programme uniquely aims to fill several gaps in the education landscape.  For example:
- Generalist data scientist Bachelors.
- Graduates with deep expertise in Data Science and a specialisation in software development (obtained for example by combining with the MSc in Software Development at ITU, to which Bachelors will have right to admission).
- Graduates with deep expertise in Data Science and a specialisation in business informatics (obtained for example by combining with the MSc in Digital Innovation and Management at ITU, to which Bachelors will have right to admission).

In addition to the two above-mentioned MSc programmes which give right to admission examples of MSc programmes where BSc in Data Science qualify for admission based on the universities’ advertised requirements.

- ITU: Games
- ITU: Digital Design and Communication
- DTU: Computer science and engineering
- AAU (Copenhagen): Operations and Innovation Management - Media Management
- AAU (Copenhagen): Operations and Innovation Management - Global Management
- DTU: Mathematical modeling and computation
- KU/AU/AAU/SDU: Computer Science
- KU: IT & Cognition

Examples of job profiles for which these education combinations prepare graduates are listed in Example Job Profiles in upload: Dokumentation af efterspørgsel på uddannelsesprofil.

Target group.  The target group for the BSc in Data Science is graduates from all high schools (Upper secondary education) in Denmark (stx, htx, hhx, hf, as well as international schools). Applicants should have strong math skills and solid English skills, combined with an interest in statistics and computing for science, social science and/or business.  ITU's Data Science program aims to be attractive to a group of students that are less likely to seek an education in a technical area, thus increasing the number of graduates with strong competences in information technology. This is because the program is designed for students with no prior programming experience and focuses on application of theoretical skills to a wide range of real-world problems.

As the programme has two tracks, the target group consists partly of potential students with a strong technical ambition and partly of potential students with an interest in the business aspect of Data Science.  Regardless of which track applicants might choose, no programming skills are required at time of application.

We expect the BSc in Data Science to be popular with a range of applicants. The program will also focus on strategic and sustainable approaches for attracting and retaining female applicants. Many female high school students in Denmark are strong in mathematics and are generally high achievers. However, current research on women in computing suggests that female students are more motivated by the potential of their scientific or technical skills to having an impact and solving the world’s problems. The BSc in Data Science is a programme heavy on mathematics and technical skills for the sole purpose of enabling Bachelors, using these skills, to effectively solve complex real world problems involving data.  ITU has also devoted resources, in collaboration with employers panels, on understanding how to increase female intake in its programmes and produce more females in IT.  Therefore, the ITU has an ambitious goal of enrolling and retaining 30% female students per year.

Specific minimum requirements:
- English, B-level with a grade point average of at least 6.
- Mathematics, A-level with a grade point average of at least 6.
However both grades are ITU standards for technical BSc programmes and in general it is expected that Data Science students achieve a higher Mathematics grade point average.

Since the programme aims to be attractive to a group of students that are less likely to seek an education in a technical area, there will be an increase in the number of graduates with strong competences in information technology. We therefore do not foresee any direct consequences for existing technical programmes. Furthermore the three current BSc programmes at ITU have many applicants and high grade point average entry cut-offs.  This summer, ITU successfully increased the number of applicants to the BSc in Software Development and admitted 50% more students without compromising the quality of the students.  In addition, not all students choosing ITU as a first choice were admitted, due to capacity limits.

2017: 50

NA