Introduction to AI
Artificial Intelligence (AI) refers to any machine or software that can learn, reason, or act in ways that we deem as intelligent. Machine Learning (ML) is an application of Al (sometimes referred to as a subset of Al approaches). ML is about using models of data to help a machine learn without being instructed, it can be thought of as self-programming. In that sense even the simplest tool like a least-squares fit or linear regression is a form of machine-learning – the algorithm ‘discovers’ how to fit data by itself.
General Introductory or Review Material
An excellent starting point for both novices and those with some existing understanding is the Elements of AI website. This includes free lesson material as well as options to sign up for more and has been developed by MinnaLearn and the University of Helsinki. You can dive right into their introductory courses or the starting points of algorithm development.
The Frontier Development Lab (FDL) is a public-private collaboration between NASA, the US Department of Energy, and the SETI Institute, with support of from AI/ML technology leaders and subject matter experts from the private sector. Notable partners include Google Cloud, NVIDIA, Lockheed Martin, Intel, the Luxembourg Space Agency, and Mayo Clinic. The FDL site has numerous resources, including publications describing projects, workshops, roadmaps and more.
Harvard University offers a free online course: Introduction to Artificial Intelligence with Python.
Microsoft offers a variety of learning materials as part of its AI Skills Initiative. Including on generative AI.
Stanford University offers a number of free video lectures on AI and Machine Learning.
An extensive but mostly high-level introduction to AI and ML from Stockholm University is available as a PDF. A similar document is available at MIT here.
A wide variety of video resources exist that range from very basic to increasingly sophisticated (including coding guides). For example, see this nearly 10-hour course on AI and ML learning.
Additional learning resources:
Basics of Machine Learning with TensorFlow. Online courses and pointers.
Startup Code Resources
A major code resource is the Keras open-source library that provides a Python interface to a variety of artificial neural net frameworks/packages, including Tensorflow and PyTorch (see more details in the links from Algorithm Central on this site). Its site includes worked examples.
Another major code resource is the scikit-learn library, with a wealth of well-organized tutorials with links to code and implementation examples. Many of these are also linked to from Algorithm Central.
There are many excellent tutorials and walkthroughs for many AI/ML applications online. For example, Jason Brownlee’s excellent Machine Learning Mastery site that includes a step-by-step resource to approaches and to getting code implemented very quickly.
Graph neural nets:
An introduction to graph neutral networks (GNN), which are a type of neutral networks that specialize in working with graph data.