Is data science more difficult than software engineering

Data Scientist: definition, job description & training

Data scientist training - With the massively increasing volume of data in companies and organizations and the associated need for data analysis, the need for specialists is increasing. A job description that is closely related and has been described by the Harvard Business Review as the "Sexiest Job of the 21st Century" is the data scientist. In this context, the question of data scientist training, opportunities for on-the-job training and suitable courses of study also arises.

Need and shortage of data scientists

Why is there a lack of data scientists on the market? The question of the deficiency should be preceded by the question of how the term data scientist is interpreted. A rough distinction can be made between two job profiles:

  • Enterprise Data Scientist: To a certain extent represents a mix of business economist, IT specialist, statistician and communication expert.
  • Academic Data Scientist: Develops pure algorithms and works with “ideal” data and is less practice-oriented than methodical-oriented.

In the enterprise environment, it seldom happens that a completely new algorithm is developed. Rather, existing concepts are adapted or expanded to the specific problem, since a completely new development of modeling processes often takes too long.

A study by McKinsey Global shows that demand in the US alone will far exceed supply in the coming year. A difficulty that is not reflected in the mere numbers: There is no such thinga job profile of the data scientist or just a special data scientist training. The requirements in the respective industries are very different. Therefore, the specific question arises: What different possibilities are there for a data scientist training in Germany and what does everyday professional life look like?

Data science definition

Data science - i.e. the science of data - is initially a bundle of different disciplines such as Computer science, mathematics, business administration and statistics. The origin of the subject is not, as one might assume, the university, but rather it developed out of the economy in the course of changing needs. Therefore, the high practical relevance of the job description as well as the science of data science and, last but not least, the data scientist training is justified.

Generally speaking, data science is about To study data using scientific methods and to be used in the context of companies and organizations. The requirement profile for a data scientist grows accordingly as his work is embedded in a company.

The job description of the data scientist is shaped by practice

Data scientists are not only concerned with evaluating data, they have toUnderstand business relationships and communicate the resultsn can. The majority of the day-to-day business of a data scientist, however, consists of identifying and compiling suitable data sources as well as preparing and carrying out the analyzes.

A data scientist bears a great deal of responsibility because of theResults of the data analysiscan depend a lot. That is why it is of enormous importance to check the underlying data again and again for plausibility, completeness, correctness and relevance.

Solve problems like a detective

The "Enterprise Data Scientist" can again be subdivided intointernal Data scientists employed by companies, andexternal Data scientists who advising are. Against the background of digitization and Industry 4.0, for example, external parties are often consulted by strategy committees.

As a service provider, you also work together with the various specialist departments in a company, create root cause analyzes on specific issues or act as “sparring partners” for internal data scientists. In this role, they have an unbiased view of the facts, can bring in fresh ideas and point out alternatives to specialist departments that they may not have thought of before. They also actively offer companies help, raise questions or first draw the attention of departments to possible solutions.

Ultimately, data scientists translate requirements into abstract data-based questions and then develop solutions that answer specific business questions. The procedure is based on hypotheses that are either rejected or confirmed. This hypothesis-driven, experimental way of working is very similar to scientific work and this explains the term data scientist.

Courage is also requiredQuestioning problems: What should be achieved and why? In the search for the solution to very tricky problems, the data scientist acts almost like a kind of top detective.

If you look at this comprehensive and very demanding requirement profile, it quickly becomes clear why there is a shortage of data scientists. The combination of very well-developed communication skills and great technical know-how is a major hurdle.

Good to know:
Gartner and McKinsey assume that the demand for data scientists will already be in 201760% bigger will be than the existing offer. IDC puts the number of data scientists required by 2018 at just over a million and sees a need for five times as much employees who can demonstrate good skills in data management and data interpretation.

In the meantime, there are (advanced) courses and further training opportunities for data scientists in many places in Germany, Switzerland and Austria. The success of these measures has yet to be proven in reality. Often there is still a gap between theory and practice. Few companies worldwide therefore also offer trainee programs for data science andTrainee-Programs for data engineers.

Technical know-how and communicative strength

The basis is a good knowledge of computer science, business administration, mathematics and statistics. A data scientist must therefore:

  • Understand business processes
  • Interpret the results of analyzes
  • Understand data-generating processes

But also thatdeep understanding of data structures, banks and models are mandatory competencies. In addition, there are programming skills to be able to work or interact with this data. This includes, among other things, the linking of different data sources, the creation of complex queries and the control of very large amounts of data.

Statistical and analytical skills come into play when predictions are made from historical datazfuture events are to be derived. The ability to understand and analyze processes or to visually process data and analysis results is also very important.

The profile is rounded off by ahigh problem-solving skills andgood communication skills. These are necessary because complex facts and models have to be communicated in such a way that management, users and customers trust the solution, and so the customer perspective and vision are not lost on the way through the data jungle. Because the point is to tell the story that lies in this data, and pack it appropriately and relevantly for each target group.

Research and study: data science at German universities

Even if data science developed out of the economy, scientific engagement with the topic is now an integral part of the university landscape in Germany. The different research areas in which data science is applied show how universally applicable data science methods are. Ranging from medicine to the humanities to space research can be found numerous research areas in which data analysis brings new knowledge.

At the same time, universities and technical colleges offer the possibility of data scientist training. More than 20 universities and colleges in Germany and Austria are now offering data science courses. The majority of these are master’s courses.

Anyone who chooses this path to become a data scientist should make sure that they have knowledge of thefollowing five areas to acquire:

  1. Analytics
  2. Data management
  3. Information design & communication
  4. entrepreneurship
  5. IT

In our article on "Artificial Intelligence and Data Science Studies" you can find more information about the courses offered for Bachelor and Master degrees.

Interest in digital professions is increasing sharply

In Germany, the job of data scientist is more popular than any other. This was also recently revealed by a data analysis by the job portal Glassdoor. Around half a million search queries were evaluated. In addition to the data scientist, who landed in first place of all inquiries, a total of 5 new, digital professions landed in the top 10. Among others, the software developer (4th place), the data analyst (8th place) and the UX Designer (9th place). The job portal Monster.de also recorded a doubling of the search queries for the profession of data scientist in the past few months.

According to a current study by the job platform “Jobfit”, there is not only a strong increase in the demand for data scientists. The analysis of more than 64,000 job advertisements has shown that, above all, an academic background is part of the standard for advertisements and that soft skills such as communication skills, teamwork and creativity are required before knowledge such as SQL or machine learning.

Data scientist training, further education and "training on the job"

The But studying is not the only wayto become a data scientist. Rather, there is a great opportunity here for engineers, economists, statisticians, mathematicians or related fields. Commercial providers such as the Fraunhofer Society offer training, courses and further education in which individual skills can be learned in a targeted manner. The advantage of these alternative training paths for data scientist training is that practical knowledge from certain areas is often already available. Those who have the appropriate qualifications can also acquire key qualifications as part of a trainee program.

Training courses for beginners or our data academy are also a good opportunity for a first approach towards a data scientist training. In addition, it is important to us at Alexander Thamm GmbH to enable our customers and the employees of our partner companies through training courses Realize the added value from your data.

Data scientist training

There are many hopes and opportunities associated with the profession of data scientist. The job titles and the training paths are in part still inconsistent in the still young field of activity, but they reflect the great variety of fields of application of data scientists. Everything is possible, from marketing to Industry 4.0.

The high degree of specialization also makes it difficult to establish or recommend uniform vocational training or uniform studies as the golden path. Because it is precisely this wide range and the high practical relevance that make the profession and the data scientist training oneattractive career option. Skilled workers and specialists from a certain area can take suitable measures to obtain further training and thus obtain a promising and currently most popular job.

There is no such thing as a perfect data scientist

Of course, specialists contribute differently to projects according to their respective strengths and advantages, and data scientists also develop their own priorities in their work. In principle, however, all applicants in the data science field require and promote all of these different skills.

Incidentally, we have not yet seen the perfect data scientist and if you take the sum of all skills and compare them to the rapid pace of technological development, there will probably never be a perfect data scientist either. Rather, it is about a data scientist forming the bracket in order to be able to solve data-driven questions from start to finish and to deploy experts such as statisticians or data engineers in a goal-oriented manner.

If, for example, the colleague from the department is dissatisfied with the "performance" of his work, then the data scientist must be able to understand the following points of view:

  1. Business perspective: For example, are my project goals at risk?
  2. Data view: Are my database queries running too slowly?
  3. Analytics perspective:Is the forecast quality / model performance too bad or is the data visualization in the display of the data too slow?

Without the data scientist, the experts here would ponder for hours what the colleague might have meant.

Conclusion:

The prerequisites for a data scientist are excellent: very good earning potential, a diverse, multi-faceted area of ​​responsibility and, above all, great future potential. Although there is still some resistance to overcome in companies - keyword "silo thinking" - and it is often a question of corporate culture whether digitization approaches will prevail, the "megatrend" digitization has long ceased to be a trend, but an exponentially accelerating development can no longer be stopped.

Intelligent machines, for example, will take over more and more human activities, including in the cognitive area, such as pattern recognition and generating ideas. The point here is not to replace people, but to complement the respective skills and processes in a meaningful way. This is precisely why the need for data scientists will continue to increase.

With their expertise, it will be possible in the future to work with intelligent machines and not against them. Because they manage to translate the business or technically oriented question into a data-driven question - and that is not possible without a well-founded evaluation and development of knowledge.

In the foreseeable future, hardly any company will be able to do without the services of data scientists, because big data and data analyzes will no longer be just "nice to have" but will be decisive for business success and competitiveness. That is why we must continue to invest intensively in the training of specialists.