Fundamentals of Machine Learning

Machine learning (ML) is a subfield of broader artificial intelligence (AI) that through programming gives computers the ability to learn from data (Alpaydin E, Introduction to machine learning. MIT Press, Cambridge, 2014; Géron A, Hands-on machine learning with Scikit-learn, Keras, and TensorFlow: concepts, tools, and techniques to build intelligent systems. O'Reilly Media, Inc., Sebastapol, 2019; Goodfellow I, Bengio Y, Courville A, Deep learning. MIT Press, Cambridge, 2016). Owning advantages in tackling problems that are too complex for explicitly programmed algorithmsAlgorithms, ML is considered as a powerful technique for making data-driven predictions and decisions. Early applications of ML, launched a few decades ago, were restricted to some specialized tasks, such as pattern recognition. With explosive growths of data and computational power over the succeeding decades, nowadays ML can be found in numerous applications even in our daily lives, from filtering email spam, ranking the search results, recommending videos and posts to language translation. The use of ML also extends widely into scientific domains, among which optics and photonics are, surprisingly or not, actively involved. In this chapter, we discuss the basics of ML in a descriptive manner. In particular, focus will be devoted to its most vibrant subfield of deep learningDeep learning (DL) (Goodfellow I, Bengio Y, Courville A, Deep learning. MIT Press, Cambridge, 2016; Kelleher JD, Deep learning. MIT Press, Cambridge, 2019).