Data Science

Turning your data into knowledge

What is Data Science?

Data science uses mathematical methods and their software-based implementation to extract value-adding correlations from large quantities of data.

Examples of such data include physical measuring points or event data from an IT system. Before identified data correlations can be successfully implemented in practice, such as in the form of smart services, they must be reinforced, interpreted and validated with the aid of domain knowledge and developed through algorithms. Procedures used to extract correlations and produce algorithms range from classic statistical methods to data exploration using information theory methods through to machine learning processes.

Data Science at a glance

Use mathematical
methods

Extract
correlations from data

Develop analysis
algorithms

Implement
smart services

Why domain knowledge is so important

Domain knowledge and a more in-depth interdisciplinary exchange with subject-matter experts is essential for the process of data analysis. On one hand, it helps ask the right questions – that is, it determines the methods that should be used to extract conclusions from the data. On the other hand, the specialist knowledge is used to interpret the analysis findings and to separate the relevant correlations from the non-relevant. Incorporating this knowledge thus helps to reduce the highly complex nature of the analysis process, which often makes it possible to solve a specific problem in the first place.

How do data scientists at Katana work?

Development of a solution strategy

To begin with, our data scientists meet with the specialists and domain experts at your company to discuss which components of your systems should and can undergo a data-related analysis. In addition, it is important to know which use cases or smart servicesyou are planning to implement with the help of big data analytics. In the next step, our data scientists will perform analyses to determine whether your data contains the information necessary to carry out your use cases. The results of the analysis are then formulated in meetings with your experts.

Algorithmic implementation of solution strategies

Based on the findings, our analysis specialists will design a solution strategy and put it into practice with the algorithms. The algorithms can recognize a wide variety of patterns in your data and provide early warning as soon as any irregularities are detected (see Anomaly Recognition). In addition, they are capable of precisely forecasting the chronological development of important parameters for your systems, which is a necessary precondition for the predictive maintenance of your machines (see Predictive Maintenance).

Validation of algorithms

The validation of algorithms is performed in close cooperation with our data scientists and your experts. Your and your team’s satisfaction with the results of our progress constitutes the basis for implementing an initial solution. In close cooperation with our data engineers and your IT experts, the data analytics solution is then embedded in your system.

Eine Herausforderung der Digitalisierung ist die Entwicklung von Geschäftsmodellen und Technologien,die eine Nutzung der Daten ermöglichen, ohne die Privatsphäre Einzelner oder die Sicherheit der Daten im Allgemeinen zu gefährden.

Quelle: Prof. Dr.-Ing. Ulrike Meyer, Professorin für IT-Sicherheit, RWTH Aachen

Any questions? We are happy to help.

Katana, the data analytics segment of the USU Group, has extensive experience in the application of machine learning methods in the industrial sector. Take advantage of our knowledge and our solutions for building your data-driven business models to reduce your costs and improve your quality and value added.