Elements of Adaptive Automatic Control

There are multiple ways to approach Artificial Intelligence in relation to human or "natural" intelligence. At the center, the core aspect of behaviour characterizes the level of efficiency and adaptiveness to the environment, or how "intelligent" this behaviour looks like, regardless if it refers to human, animal or machine.
In this lecture, concepts of AI and Machine Learning (ML) are investigated in the context of Control Theory, i.e., designing automatic control for various tasks. The classic PID controllers are just a small family of design options for linear systems, while in Signal Processing the analogous approaches usually evolve around linear models like ARMA and ARIMA, or equivalent designs in the frequency domain. However, in real-world applications the "linearization" option is valid only for short time frames and/or high-order approximate linear models. AI and ML provide the means to extend these tools towards robust, adaptive designs for reactive and predictive control systems, which are ubiquitus today in the car industry, industrial processes, civil aviation, space technologies, etc.

Keywords: Machine Learning, Data Analytics, AI, Artificial Intelligence, lecture, Automatic Control, Control Theory, Adaptive Systems, Signal Processing
Video: https://youtu.be/HgnbdU3UiY0