Machine learning: a personal tour

Machine learning is about learning patterns in data. More specifically, it's about ‘input-output’ paradigms and computational frameworks, often, but not always applied to inputs to estimate outputs. It's also about the practice of abstraction and representation of of data, about model construction and estimation and calibrations; about testing, and assessing a model's performance and, last but not least, the application and interpretation of a model. Without touching every aspect, we'll describe the main thrust of machine learning, from the foundations up to applications and point out how machine learning is used in practice.