The data scientists of the future

In the fast development of technology, DSV has together with the Department of Statistics, developed a new master’s programme, aiming to meet the demand for future data scientists.

This program takes a new approach by combining three subjects: data science, statistics and decision analysis.

Ioanna Miliou, one of the programme coordinators and creators, talks about the inspiration for the courses’ design, its structure and the thoughts behind the programme.

Ioanna Miliou Photo: Carina Bergholm
Ioanna Miliou programme coordinator for the new master’s Program in data science, statistics and decision analysis at DSV. Photo: Carina Bergholm

“We want to create the data scientists of the future. And not in the way that it has been done before. A good data scientist needs a robust statistical background. Not necessarily with complex maths, but a solid foundation in statistics. The students will also gain knowledge about decision problems and various types of risk analysis and learn how to implement it in a corporate setting”, says Ioanna Miliou.

We want to create the data scientists of the future. And not in the way that it has been done before

The collaboration between the departments emerged from a recognition of gaps in the previous master’s programme in data science and decision analysis.

“Statistics is the base of machine learning and data mining algorithms today. Therefore, it is great if we can create data scientists – or if you speak more generally, computer scientists – that now have this part incorporated in their master’s degree.”

Course structure

The program's structure consists of the three pillars – data science, statistics and decision analysis – with 33 percent each of the curriculum.

“We are designing the content of the courses to cross over and bring one field inside the other. There will be a specialist course leader from each field for each course but we want to create a flow between the courses.”

“We will guest lecture in each other’s courses and have similar paradigms throughout the courses. The students will feel how things go together.”

We want to create a flow between the courses

Practical experience

Most of the courses will include both traditional style lectures and also lab work, while some courses will have projects as well.

“Almost 50 percent will consist of practical testing to implement the things the students have learnt in class. All courses in data science have got a programming part. Everything they learn in class, they will learn to implement in coding. We will be discussing the theoretical part and, sometimes, the maths behind it. The students then go to the labs and “make it work” with real data. In data science we use Python and in statistics we use R. They are very similar languages. It is not a hard transition from one language to the other. They are both very intuitive programming languages.”

Accessibility for all backgrounds

The program is accessible for students with only a little bit of programming skills. The courses are designed as a continuum, gradually increasing in difficulty.

“We will start with a course in foundations of data science, with easy programming assignments and study the theory at a lower level. After that students will be prepared for the data mining course. Where the students will see the same type of algorithms and assignments but more complicated. In the spring, we will go in to deep learning with the most complex algorithms. Student will be able to follow the courses building on each other.”

Text: Carina Bergholm