Author: Diego Gosmar

Publisher:

Published: 2020-08-23

Total Pages: 258

ISBN-13:

★ COLOR VERSION ★ New edition updated 2021! Machine learning is one of the most powerful artificial intelligence techniques, capable of efficiently managing and analyzing large amounts of data, to provide accurate predictions, automated decisions and deliver unprecedented business benefits. This volume aims to illustrate in the simplest possible way which are the main approaches in the Machine Learning universe, as well as providing some examples of real applications from which the reader can draw inspiration to understand the benefits and design applications of common interest. Among the covered topics you will find: * Practical applications: regression and classification predictions * Sentiment analysis * Speech Analytics * Image recognition * Performance analysis * Numerous examples and graphical displays of the results * Wavelet Transform for AI non-stationary signal processing * Supervised, Unsupervised and Reinforcement Learning * Artificial Neural Networks, Convolutional Neural Networks, Recurrent Neural Networks * AutoML * MLOps and Pipeline for Distributed Architectures to improve Governance and Scalability The author describes the essential principles and methods of ML (Machine Learning) clearly, making the book suitable even for non-IT readers or data scientists who are experts in the field. Business innovation managers and departments can also benefit from reading this book to better understand how ML can streamline its operations and increase productivity, with an eye to the future. After a first introduction to the concepts of data science and the nomenclature often adopted when it comes to Machine Learning, the book offers a description of the three main methodologies adopted today, trying to analyze both the benefits and the critical issues. Some of the most common learning models are illustrated and the various steps for preparing the data are then analyzed together with the training, testing and accuracy assessment phases. Some of the IT tools that can be used to work on Machine Learning are then described (with emphasis on Open Source ones). The second part of the book deals with different techniques of Regression, Classification and Deep Learning, as well as the methodologies to optimize the results and combine the adopted algorithms. We examine the subject of model interpretability and also of AI security, to move on to an overview of visualization and analysis techniques during Machine Learning processes. The final part focuses on real applications. Two practical cases related to real business applications are dealt with, the approaches to face them, the tools adopted are described and all the source code is made available, commenting it step by step for greater understanding. This volume tries to deal with concepts related to the world of Machine Learning using a language suitable for a wider audience possible, because Machine Learning is part of the fascinating vast world of data science, which brings together various skills: technology, analysis and business understanding.