Data is everywhere. The trick is to extract important connections and trends from this. The use of statistical methods has broadened considerably in recent years and the importance of their application has increased enormously. Not only in classic environments such as laboratories and industry, but in all technical fields. Expert Koo Rijpkema, senior lecturer in applied statistics (TU/e) and PAOTM course leader, knows how important data analysis is and how to apply it successfully. We spoke to him!
Why is data analysis an indispensable discipline in a technical environment?
“To measure is to know” is the starting point. Efficiently setting up experiments, assessing the quality of the obtained data and developing suitable decision support models has an important role. Just like the effective reporting and communication of results and conclusions. And finding your way through the multitude of automatically logged data in industrial processes. Many methods and (software) tools are available for this nowadays. But because these are constantly being developed, for responsible use it is important to have and maintain an overview and insight into the available options and the do's and don'ts. It is also important to be able to critically assess their value on data-based reports and advice.
How fast are the developments in the field?
Fast! Public-domain available data analysis software, such as Python and R, can immediately respond to new possibilities and trends. With R, for example, the number of available packages approaches 20,000, while in Python the number of packages is also steadily increasing. There are also more and more initiatives to link the possibilities of R and Python and to achieve a best of both worlds!
Can you give an example of a technical field where data analysis is rapidly gaining ground?
This happens in many technical fields. Think of energy, the electricity supply and 'the smart grid' and infrastructure that will be needed in the future for charging electric vehicles. In recent years, for example from the water boards, there has been a great deal of interest in water management in forecasting methods to map the consequences of climate change for the risks of flooding or drought. PAOTM has also dedicated special courses to this subject! You can also think of Autonomous driving, Model predictive maintenance, pharmaceutical industries and High-Throughput Experimentation (HTE) in chemistry.
When data analysis is applied incorrectly, this can have major (social) consequences. How do you prevent that?
Big data, if misapplied, can have devastating consequences. The reasons can be anything like lack of tools, technical issues, low quality data, wrong or unnecessary data. One of the most famous big data mistakes was Google Flu Trends. This web service was supposed to predict flu outbreaks in about 25 countries. The logic was simple: analyze Google searches about flu in a particular region. The results turned out to deviate no less than 140 percent from reality! The algorithm turned out to be flawed and the search terms were chosen too broadly. And what about sensitive information? Then the consequences are really unacceptable: for example when targeting a target group where private data becomes known (eg the stores Target and OfficeMax). And so there are many examples of wrong data usage. Fortunately, there are remedies for this. We pay ample attention to this in the PAOTM courses by not only discussing the methods and techniques, but also the context, use and application.
Are the courses only for data analysts? Or can you, as a manager, also learn how to read and interpret data?
There are not only courses for specialists, who, for example, carry out data analysis projects themselves, but also for generalists who manage such projects or who want to make responsible decisions based on reports. The Practical Data Science and Essentials of Predictive Analytics courses are particularly suitable for this, because interpretation and substantiation of the conclusions found play a central role! They also bridge the gap between the more classical statistical data analysis and the modern data science field.
PAO Techniek en Management offers a series of courses in which you gain broad basic statistical knowledge or in which you can opt for further in-depth knowledge in the more specialist courses.
Regardless of which (technical) field of work you are active in. Koo will inform you fully in a few days. In all courses you learn to work with representative public domain available software, such as R or Python, and you can bring your own cases.
Register quickly for the upcoming courses in November and January!
Practical data science - from November 24, 2022
Time series analysis and forecasting - from November 22, 2022
Multivariate data analysis - from January 11, 2023
We also offer the tailor made incompany courses Essentials of predictive analytics, Design of experiments and Data mining & business analytics.
Want to see a compact overview of all courses?
View our data analysis and applied statistics webpage.