In the Mobility Insights team of Swisscom, we are building algorithms that allow us to quantify collective mobility at the scale of the country. Quantifying mobility on such a massive scale offers unprecedented promises to improve infrastructures and make the country run better.
The goal of this project is to enrich our Mobility Platform by integrating a Machine Learning component into our feedback platform. This (1) allows a user to associate her mobility traces (trajectories) with labels (trips, mode of transport, etc) and (2) integrate a machine learning algorithms that automatically infer these labels.
The student would be responsible of investigating and implementing machine learning algorithms that are able to automatically annotate these trajectories. This requires both an understanding of Machine Learning but also extensive Software Engineering skills
- Real-word application of Big data: We deal with more than 1 million network events per second
- Gain experience in implementing the full machine learning pipeline: from labeling, to training and classification.
- You will be working with the team, taking an iterative approach that involves writing production, or near-production, code from day one
- Improvement of coding and theoretical skills.
- A hands-on data-science project, with a positive societal impact.
- Your work will be highly visible, with a demo, an article and a video on Linkedin and on our research portal, like https://research.swisscom.ai/blog/augmenting-ai-beyond-its-own-boundaries/ (video embedded). It's a door opener towards a good position afterwards.
- We highly support publishing and aim for top conferences.
- Strong understanding of machine learning.
- Strong proficiency in software engineering (Python and/or Scala).
- Experience in UIs is a plus.
Platform to label mobility data and improve the quality of our MIP