ZOMA, the Zoo Ontology for Machine Algorithms, houses information about AI/ML theorems, algorithms, and frameworks in order to provide a one-stop resource for exploring these objects.
If you want to dive right in, here's how to do so. You can visit the zoo itself, take a look at the xml file that houses the data, or use the generator to create a new structured entry. If you make something cool, then by all means email it to me so I can add it! And lastly, check this out if you would like a crash-course in machine learning algorithms.
As these are mathematic objects that can be implemented as computer code, they cannot physically reside anywhere, and there may not be a canonical or standard coded implementation for any of these. So a Zoo, a special type of information organization that draws on features unique to Galleries, Libraries, Archives, and Museums (GLAMs) solves this problem.
Let's face it. Machine Learning is a wild frontier. Neural Nets. Big Data. Data Science. Really Big Data Science -- these terms draw massive amounts of hype. That hype is largely disproportionate to what is common knowledge about what these algorithms are and can/cannot do.
If there were only a standard format for talking about these as if they were actual objects or organisms the way we can say a painting is Pre-Raphaelite or a bacteria is gram-negative. Algorithms, despite being abstract objects, have concrete properties such as their runtime complexity, and have a certain taxonomy. Neural Networks, for example, are incredibly diverse as they are composed from smaller neural units.
ZOMA works in two ways. It is a curated repository (a Zoo) that houses "tamed" algorithms. These algorithms are "tamed" by converting them into XML, forming a linkable datastructure. Down the line, this datastructure will allow people to make queries (and even use ML on ML itself) regarding how fast or slow an algorithm runs, what it is composed from, and so on.
This work is licensed under a Creative Commons Attribution 4.0 International License.