The classic book on artificial associative memory is T. Kohonen’s “Self-Organization and Associative Memory” [1]. Kohonen’s most famous contribution is the Self-Organizing Map. Prof. Kohonen recently passed away on December 13, 2021, at the age of 87.
There is also a very informative review article on associative memories at The Gradient.
“Artificial neural networks and deep learning have taken center stage as the tools of choice for many contemporary machine learning practitioners and researchers. But there are many cases where you need something more powerful than basic statistical analysis, yet not as complex or compute-intensive as a deep neural network. History provides us with many approaches dating to the Multilayer Perceptron era. Yet many of these alternate methods languish in the shadows despite modern advances in enabling technologies. One such approach is called the associative memory.” [2]
“In fact, most ML textbooks don’t even make mention of associative memories, which I personally feel is a grave oversight (I confirmed that my trusty copy of Mitchell’s venerable Machine Learning sheds no light on these algorithms). So the next time you need a robust pattern matching system, noise-resistant pattern recaller, bidirectional learning system, or have a few-shot requirement, consider adding associative memory approaches to your ML toolbox and you might be surprised how far they can take you.” [2]
References
[1] Teuvo Kohonen, “Self-Organization and Associative Memory”, Springer-Verlag, 1984
[2] Robert Bates, “Don’t Forget About Associative Memories“, The Gradient, 2020.
Suresh Emre
Renaissance Universal 