The creativity of a Data Scientist
Think of a creative baker: the image that pops into your mind is probably of someone with an endless stream of ideas who then translates them into beautiful (and hopefully tasty) confectionery. A creative architect? Of course: someone who takes a seemingly boring project and brings it to life, delighting the people who occupy and move around in a new space.
What about a data scientist? Does creativity in the field come in the form of developing new machine learning models? Applying academic research to real-world problems? Running innovative A/B tests? Finding unlikely solutions to thorny problems? In a field so resistant to stable definition, it can be hard to say for sure; but when we see it, we know it's there, and happily so. And here are just a few examples.
- Learn howow graphical analysis can help detect healthcare fraud. "What happens when the problem is less about the number and more about the network?" Lina Faik combines machine learning and graphical analysis to explore a specific use case - fraud detection in the healthcare sector - and shows how it can improve model performance without sacrificing interpretability.
- Avoid a common pitfall to make sense of your data. Eric J. Daza, comes down to calling your shot after you've made it, in other words, adapting your theory to the results you have. It's a tempting move in the face of uncertain or confusing results, but it's a trap you want to stay aware of, or statistical overlearning awaits you.
- Improve your storytelling and presentation skills. Vicky Yu has covered a lot of data topics, including thoughts on the changing role of data analysts. However, whatever the job description, stakeholders can't appreciate your contributions if you don't explain them in a way they can understand.
- Explore new frontiers at the intersection of data and biology. Daniel Bojar's latest work focuses on glycans, complex carbohydrates that play a key role in processes such as viral infection and tumor immune evasion. Here, he turns to convolutional neural networks (CNGs) to define a new state-of-the-art for the analysis of these compounds.
- Discover how a data team developed a new approach to meet a business need. Working in industry requires fast, creative solutions to complex problems. Jonathan Ramkissoon, explains the whole process that led him to leverage semantic understanding and user behavior to classify the types of documents users upload to the platform.
- Break certain rules (or create new ones). Creativity often lies in the balance between mastering the rules and bending them thoughtfully and intentionally. Semi-Koen has compiled no less than 35 pieces of conventional software development rules that also apply to data science. Many are ironic, but all are an invitation to rethink the way we conduct our work.
To discover all the skills a Data Scientist needs to have, read our article on Data Science skills!
Looking for a new opportunity in Big Data? Click here for a sneak preview of our new AI-powered job platform.
Thanks for reading! If you'd like to read our next articles on Data Science, you can follow us on Medium, Facebook, LinkedIn and Twitter to be notified when a new article is published!