IFEES // Global Webinar Machine Learning from Data to Knowledge
Machine Learning is an important part of what is commonly called Artificial Intelligence, and it is a mature and well established discipline intended to provide computers with an ability to take decisions without explicitly programming any established rules or concepts, but simply by using the available data in order to infer or extract knowledge from it. The independence of machine learning strategies from any necessity to provide them with prior knowledge about the observed phenomena is one of the keys that made these techniques so powerful and hence they are making an impact in our everyday activities. We can talk about three different types of learning machines from a structural point of view. Kernel machines and ensemble machines, which started to become popular in the late 90’s, and deep learning machines, which are rooted in the good old neural networks (1980) but that became popular in the early 2000 thanks to theoretical advances and the popularization and improvement of supercomputers and parallel computing both from the software and hardware points of view. Besides, other structures and approaches can be seen in reinforcement learning. In this 30 minute presentation, we will overview these techniques and main features, and we will discuss where they are used depending on the problem that we have at hand, and the quantity and characteristics of the available data. Manel Martínez-Ramón earned his degree in Telecommunications Engineering at Universitat Politècnica de Catalunya in 1996 and a PhD in Telecommunications Technologies at Universidad Carlos III de Madrid in 1999. Since then he has been conducting research in several areas in Machine Learning applied to communications, medicine, electrical grid and others. His scholarly contributions include theoretical contributions in kernel learning, support vector machines and gaussian processes, where, together with his coauthors, he introduced these techniques in signal processing and communications (Rojo, Martínez, Muñoz, Camps, “Digital Signal Processing with kernel Methods”, Wiley, 2018) and electromagnetics (Martinez, Rojo, Gupta, Christodoulou, Machine Learning Applications in Electromagnetics, Artech House, to appear). He is currently a full professor with the Department of Electrical and Computer engineering of the University of New Mexico, where he holds the King Felipe VI Endowed chair in Information Technologies. His current research is in machine learning applications to smart grid and to scientific article accelerators.
Machine Learning is an important part of what is commonly called Artificial Intelligence, and it is a mature and well established discipline intended to provide computers with an ability to take decisions without explicitly programming any established rules or concepts, but simply by using the available data in order to infer or extract knowledge from it. The independence of machine learning strategies from any necessity to provide them with prior knowledge about the observed phenomena is one of the keys that made these techniques so powerful and hence they are making an impact in our everyday activities. We can talk about three different types of learning machines from a structural point of view. Kernel machines and ensemble machines, which started to become popular in the late 90’s, and deep learning machines, which are rooted in the good old neural networks (1980) but that became popular in the early 2000 thanks to theoretical advances and the popularization and improvement of supercomputers and parallel computing both from the software and hardware points of view. Besides, other structures and approaches can be seen in reinforcement learning. In this 30 minute presentation, we will overview these techniques and main features, and we will discuss where they are used depending on the problem that we have at hand, and the quantity and characteristics of the available data. Manel Martínez-Ramón earned his degree in Telecommunications Engineering at Universitat Politècnica de Catalunya in 1996 and a PhD in Telecommunications Technologies at Universidad Carlos III de Madrid in 1999. Since then he has been conducting research in several areas in Machine Learning applied to communications, medicine, electrical grid and others. His scholarly contributions include theoretical contributions in kernel learning, support vector machines and gaussian processes, where, together with his coauthors, he introduced these techniques in signal processing and communications (Rojo, Martínez, Muñoz, Camps, “Digital Signal Processing with kernel Methods”, Wiley, 2018) and electromagnetics (Martinez, Rojo, Gupta, Christodoulou, Machine Learning Applications in Electromagnetics, Artech House, to appear). He is currently a full professor with the Department of Electrical and Computer engineering of the University of New Mexico, where he holds the King Felipe VI Endowed chair in Information Technologies. His current research is in machine learning applications to smart grid and to scientific article accelerators.