Being an entrepreneur may come with a lot of uncertainty, and many people do not want to take risks when it comes to making a living. Those usually choose the safe option and become, for instance, doctors. In semester 2, we visited an emergency medicine institute in Munich. But don’t worry, we have not changed the profession – we went there to kick off our Predictive Analytics project!
Quick Reminder: What is Predictive Analytics?
Predictive Analytics is the use of data to make better business decisions by predicting complex economic relations. The trend can be observed in almost every industry. For example, Netflix uses Algorithms to predict what customers will watch, or industrial companies predict machine failures before they happen.
Master students solving real business problems with data mining
If you want to study “something with digital technologies”, there is no way around Big Data. In our program, an entire module is assigned to cover this hot topic. Prof. Günzel, who is not only dean of the business faculty but also our program head, coordinates the module “Empirical Research Methods and Effective Data Analysis”. And because predictive analytics is something very case specific, he wants us to work on a real business case. For this, an organization provides us with a real data set, and our task is to explore and analyze it.
You’re curious what kind of data we got? Wait no longer – here comes the answer!
Partnering up with the Institute for Emergency Medicine of LMU München
Prof. Günzel chose a field where predicting events could actually save lives: Our partner would be an emergency institute responsible for planning the ambulance fleet of 11 hospitals in Munich. The data set contained patient records described by several attributes, e.g.
- the hospital at which the patient showed up
- driving time to the hospital
- a number of demographic features
- which treatment the patient received (station or ambulant)
During the kick-off meeting at the institute, we learned, for example, how patient and ambulance car data is already used to predict how many ambulance cars need to be available at each day of the week and time of the day. After the kick-off meeting, we formed groups of two. Each team chose one of the Data Mining topics Prof. Günzel had prepared, e.g. hierarchical clustering or outlier detection.
Our field trip ended with a tour through the simulation center for emergencies in the institute’s basement, and you wouldn’t guess what we discovered there…
Tour of the emergency institute: A helicopter in the basement!?
As part of a medical simulation of emergency cases, the institute has several shock rooms downstairs. The rooms follow the usual process of an accident: Room 1 simulates a car accident, room 2 the transportation to a hospital, and room 3 the emergency room. In order to train medical staff in scenarios as close to reality as possible, the institute uses real equipment: Real medical instruments, a real car, and – a real helicopter.
You might ask, how the heck did the helicopter get into the basement? Marc Lazarovici, one of our partners at the institute, explains:
Check out our impressions of the emergency simulation centre:
Student teams and Data Mining Algorithms
Let’s fast-forward to the final semester presentation, where we presented our results. Since we were given a lot of freedom to explore the data set and our topics, it was interesting to see a wide variety of results. Some groups went deep into theory, explaining how the algorithms compute results. Others focused on finding a real business problem: After obtaining an in-depth understanding of the process a patient goes through when arriving at a hospital, they were able to give suggestions on how to improve the decision making process. Another group had even studied programming in R in order to apply their algorithm to the data set.
To me, it was truly eye-opening what can be done with predictive data analysis. Saving lives was certainly not on my radar when we started this semester. Aside from the real case, we learned how big data can drive productivity. If you would also like to learn about the various applications of data analytics to help businesses thrive, join our Master program!