I collected my first machine learning experiences in 2009 with my diploma thesis. I created a recommendation system for music bands, records and tracks as collaborative filtering for the social networking site metalcon – which I have created as a startup in the past. Already in that time I taught my first class about machine learning techniques with the German Summer Highschool (Deutsche Schüler Akademie)
Most recently I am supporting the German media company Merkurist. The goal is to predict which article requests that are suggested by the users as Snips should be actually written by the journalists since they are likely to have a great performance.
I have supervised several master thesis in which students of mine used machine learning techniques. For example to automatically extract citations from scientific papers using conditional random fields or to detect bad communication in deletion discussions of wikipedia with the help of the software kit rapid miner.
With modern machine learning techniques one can do much more like predict risks, future revenue, events or even the behavior of your customers. Sounds good but you need help in your implementation? I can support you from the planning phase in which you will understand what predictive tasks you could solve with your data all the way up to a full implementation of your machine learning models.