Machine learning is used to design and implement software systems that learn from experience. Such systems are not directly programmed to solve a problem based on a fixed algorithm, but instead they further develop their own algorithm based on examples of how they should behave. This requires some form of trial and error experience, trying to solve the problem.
Therefore machine learning is very promising for problems that are solved many times, which includes many decision problems in business and economics, as well as many problems in engineering and sciences. Machine learning has proven itself through many applications in economics and life sciences, and its use has spread towards almost all fields where a class of problems I solved over and over.
Learning performance of any machine learning approach is very dependent on the type of trial and error experience, therefore researchers not only develop new algorithms but also focus on evaluating which algorithms are best used in which circumstances.
This course will focus on “a core set” of machine learning methods that have proven valuable and successful in many practical applications. This course will contrast these methods, with the aim of explaining the circumstances under which each is most appropriate.