It has become common knowledge that Data Science is one of the best skillsets you can develop in order to pursue a successful career. However, becoming a data scientist is not only about analysing data, but also about acquiring that data and communicating the results of your analysis.
To become a data scientist, first, you need a solid foundation in math and physics, and understand how machine learning algorithms are derived from basic math. You also need to be curious about new mathematical tools and emerging areas, such as deep learning and network science.
Second, you need programming and hacking skills. Most data you’ll get your hands on is raw and unstructured, and needs a lot of parsing work before you can even begin analysing it. For this purpose, it is useful to understand data mining tools, such as browser automation drivers, regular expressions, and natural language processing.
Last but not least, you need to know how to communicate results to people who have little knowledge of data science. Data visualisation tools, a good sense of design, storytelling abilities, and domain expertise are essential to carry out this task.
At the Center for Digital Education (CEDE), we provide students with the opportunity to develop all these skills, using real datasets and addressing real problems we encounter at EPFL. Our projects are challenging but exciting, and we usually reward our best students by publishing their results and showing them to the EPFL leadership.