Machine Learning is perhaps the fastest-growing field of Artificial Intelligence, as in recent years, especially after the advent of Deep Learning, it has provided a plethora of methods with very good to impressive results in almost all applications that require intelligence.

This book systematically describes the three main types of learning: supervised learning, unsupervised learning, and reinforcement learning. For each type of learning, the most important models are analyzed, such as neural networks, support vector machines (SVM), Bayesian probabilistic models, graphical models, stochastic models like the hidden Markov model (HMM), recursive models like LSTM, and many others.

Specifically for neural networks, which are a very important part of machine learning methods, a systematic and analytical presentation is provided, starting from the simple Perceptron model of a single neuron and reaching the more complex models, such as deep neural networks. For each model, the necessary mathematical background is given for understanding its operation, with only basic knowledge of probability theory and linear algebra as prerequisites.

Additionally, emphasis is placed on the algorithmic dimension of the models, as most of them are accompanied by the relevant pseudocode and application examples.

The applications of machine learning models constitute an important part of the book, given that they represent a fundamental motivation for the study and development of the models. Various applications are described, such as pattern recognition, signal and image processing, speech processing, information compression, strategy development in games, etc.

The authors: Dr. Konstantinos Diamantaras is a professor in the Department of Computer Engineering and Electronic Systems at the International Hellenic University. He is an active member of the Institute of Electrical and Electronics Engineers (IEEE) and has many years of experience in research and teaching the subject of Machine Learning, both at the Thessaloniki Technological Educational Institute and at the International Hellenic University.

Dr. Dimitris Botsis works in a technical-studies office and is an associate professor in the Department of Topography and Geoinformatics at the International Hellenic University. As part of his doctoral dissertation, he worked on the development and application of machine learning methods in hydrology and time series simulation. He continues to be actively involved in research in the field of machine learning, focusing on the latest methods and a wider range of applications.