Data Science for Starters: How to Train and Be Trained

Given the right tool, it may take only a few hours to familiarize outsiders with data science. Data science, machine learning, and artificial intelligence are drivers of change in all fields of science, including biomedicine. The computational approaches that can sip through vast collections of data, extract interesting patterns, and devise predictive models are becoming omnipresent. But only a few professionals understand the essential concepts behind data science, and even fewer engage in building models using their data. We here report on the contents of the tutorial at AIME-2022, where we aim to explain how anybody who can spare a few hours can learn the essential mechanics behind data science and machine learning. With the training we have designed, the professionals can gain enough intuition about data science to recognize opportunities that this field can offer and actively engage in data science projects. Besides good mentors and an encouraging working environment, the right tool is critical for such training. We advocate the workflow-based construction of analytical pipelines with interactive visualizations and show that they can be the key to the simplicity of the interface and flexibility to adopt analytics to any data type and problem domain.