Resources for working with Machine Learning in F# [closed]

Answer 1

In addition to what Tomas mentioned, I spent some time with Infer.NET about a year ago and found it was pretty good for continuous graphical models. I know it's improved quite a lot over the last year in both the scope of the library and F# support. I suggest checking it out and seeing if it has what you need.

Answer 2

APress has a book in "Alpha" slated for release soon: Machine Learning Projects for .NET Developers. http://www.apress.com/9781430267676

The currently existing content seems to be introductory, but quite good to learn from, and its code samples are primarily F#.

Answer 3

There isn't a single place to look for resources on F# and machine learning, but here are a couple of links that may be quite useful:

• Numerical Computing section on MSDN is a good resource on using various numerical libraries from F#. The most advanced library that implements linear algebra and other algorithsm useful in machine learning is Math.NET Numerics.

• Visualizing Data section on MSDN has some resources on charting in F#. The FSharpChart library is now maintained by Carl Nolan who regularly posts updates to his blog.

There are also a few personal pages of people who are working on relevant topics:

• Jurgen van Gael (who did PhD in machine learning) contributed to the Math.NET library and you can read about his experience here.

• Yin Zhu who wrote the Numerical Computing chapter on MSDN (and is a PhD student interested in machine learning) has quite a few excellent articles on his blog.

Answer 4

Hal Daume has implemented a lot of machine learning algorithms in OCaml and Haskell. Details see my answer in Machine learning in OCaml or Haskell?

As side from the Numerical Computing in F# book chapter on MSDN, I'd also like to recommend my Wrapper for Weka, WekaSharper. It allows you to call machine learning algorithms in Weka using an F#-friendly interface.

I wrote an article, Why F# is the language for data mining, which reflects my thinking when I finished writing a alpha/prototype-like data mining package in F#. libml is available online. But the code was written about two years ago when I started to use F#, and I didn't have time to maintain it since then.