Machine Learning by David L. Richards

• TenaciousG - a graphs tool with persistence.
• Sirb - statistics and erb. A bit of a hack -- it monkey patches other classes. It's kind of cool
• TeguGears - Tegu is a never finished gem for machine learning, all in one place. TeguGears is meant to be its core functionality.

statisticus - an immature "R" type stats package for Ruby. Don't use it -- it will break -- if you can even get it to run. He's still cleaning it up, and feels that it has potential. rbtree

A lot of machine learning uses graphs.

Approaches

• Ruberl
• Rocaml
• Blurred Weasel

Functional style programming has three primary advantages:

• Composable
• Concurrent
• Distributed
• Etc.

Fibers. They're a "new to Ruby" idea -- coroutines. They start "paused" and they are "resumed".

• PipelineElement - stolen from Pragmatic Dave (Thomas).
• NeverBlock is an I/O oriented library. Works with Fibers.