Hey everyone, my name is Job Verschuur a student of a master Applied Physics at the University of Twente in the Netherlands. I have joined the Innovation Sprint team in April for an internship of 3 months, where I will be working in the machine learning group. I am born and raised in Hengelo (ov) (the Netherlands). In addition to my studies, I play volleyball (before corona) and I like to do woodworking. My interest in woodworking started early, by making model airplanes from balsa wood.
I started at the University of Twente in my bachelor’s and followed the standard curriculum. After the bachelor’s, I started with the associated master’s, which consists of three main directions at the University of Twente: Optics, Material physics, or Fluid physics. Since I could not choose, I did most of the courses of all three directions and add some courses of the mathematical master. In the end, I did my master’s assignment in the fluid physics direction. However, I still had to do my internship…
As I mentioned before, my background is Physics and a little math. So why Innovation Sprint, which lies in a more clinical direction? As I was looking for an internship, which is quite difficult in this period due to Covid-19, I got in contact with Miriam Cabrita, a researcher of Innovation Sprint. After a short meeting, she brought me in contact with Aristodemos Pnevmatikakis, the R&D Director of Innovation Sprint. Aristodemos gave me a short introduction to the company and offered me several possible directions for my internship assignment inside the company. One of these directions, an assignment involving machine learning, got my interest since I like to play with code and because machine learning is an upcoming topic in many fields, so getting machine learning experience at a company may be helpful for my later career.
During my internship, I will be working in the machine learning group that builds Healthentia, an eClinical platform for collecting, understanding, and acting upon Real-World Data (RWD). The work I am involved in is the phenotyping of the participants. I will start with attempting to implement more clustering algorithms for the clustering of the participants in phenotypes (synthetic data). In particular, a clustering implementation with deep autoencoders. After that, the plan is to explore the modelling of the phenotypes and the possible improvements of the existing code to multidimensional modelling of attributes. Finally, I will work together with the ML team to make the associations with phenotypes online in the Healthentia platform.
I expect to learn a lot at Innovation Sprint. About machine learning, but also about the inner workings of a startup company. Furthermore, I hope to be a meaningful addition to the team during my internship.
