How to build usable machine learning solutions based on unmarked data. Many industry experts consider unsupervised learning to be another frontier in the field of artificial intelligence, which may be the key to full artificial intelligence. Since most data in the world are unmarked, conventional supervised learning cannot be applied to them. In turn, unsupervised learning can be applied to unmarked data sets in order to discover important patterns hidden deep in those data that can be almost impossible for man to discover. Author Ankur Patel shows how to use unsupervised learning using two simple Python platforms: Stukit-learning and TensorFlow (with Keras). With the included code and practical examples, data analysts will be able to identify hard to find patterns in the data and discover in-depth business relationships, detect anomalies, perform automatic variable selection and generate synthetic data sets. Simply know programming and a little experience in machine learning to address:Comparing the strengths and weaknesses of different approaches to machine learning: supervised, unsupervised and reinforced learning. Preparation and management of machine learning projects. Building an anomaly detection system to capture credit card fraud. Separating users into separate and homogeneous groups. Conducting semi-supervised learning. Development of video recommendation systems using limited Boltzmann slots. Generating synthetic images using generating antagonist networks. Researchers, engineers and students will appreciate this book full of practical techniques of unsupervised learning, written in a simple language with simple examples in Python that can be implemented quickly and effectively. Sarah NagyThe main data analyst at EdisonAnkur A. Patel is vice president of analytical computing at 7Park Data, supported by Vista Equity Partners. At 7Park Data, Ankur and his data analysis team use alternative data to develop data-related products for hedge funds and corporations and develop machine learning services for corporate customers.