ARTIFICIAL INTELLIGENCE ALGORITHMS FOR UNSUPERVISED LEARNING: CLUSTERING AND PATTERN RECOGNITION WITH NEURAL NETWORKS. Examples with MATLAB

Artificial Intelligence combines mathematical algorithms and techniques from Machine Learning, Deep Learning and Big Data to extract the knowledge contained in the data and present it in an understandable and automatic way. Neural networks and their applications are a fundamental tool to develop work in Artificial Intelligence. On the other hand, unsupervised learning is more closely aligned with Artificial Intelligence as it gives the idea that a machine can learn to identify complex processes and patterns without the need for a human to provide guidance and supervision throughout the learning process. Some examples of unsupervised learning algorithms include clustering and association rules. In the case of this type of learning, there is no pre-training data set; the problem is approached blindly and only with logical operations to guide it. Although at first glance it seems impossible, it is about the ability to solve complex problems using only input data and logical algorithms. This avoids the use of reference data. Unsupervised learning algorithms are used to discover hidden patterns in unlabeled data. Unlike supervised learning algorithms, where there is prior knowledge of the desired answers, these algorithms do not have a set of ordered data. They are responsible for determining the most important common characteristics of a group of information and then grouping them according to their similarities. Among the most interesting models are the neural networks. MATLAB implementrs the Deep Learning Toolbox specialized in the techniques of analytics based on neural networks. Throughout this book the techniques of analytics for clustering and classification based on neural networks are developed using MATLAB software