Open-Source Machine Learning Library Available

Ridgelift writes "CNet has this article on Intel's OpenML (press release). "The chipmaker's research arm on Monday said it is offering the Open Source Machine Learning Library, a collection of software that can help computers learn from various experiences. It will offer the library to interested parties for free via the Web." Open the pod bay doors, HAL."

umm, borken links?

[Anonymous Coward]

Did anybody else have success downloading stuff? The sourceforge page doesn't have shit: it's completely empty. Even some of the pdfs on the intel site are broken links.

Should of expected as much from those dumbfucks at intel.

Namespace collision?

[CableModemSniper]

http://www.khronos.org/openml/ [khronos.org]

Re:Namespace collision?

[DeadMeat (TM)]

Intel's problem is that they use too few buzzwords. Now if they'd just called it "FreeOpenML XP Extreme Edition", they'd be fine.

Re:Namespace collision?

[CableModemSniper]

I dunno Hyperthreading seems pretty buzzy to me.

LingPipe

[JohnFluxx]

Also going open source is LingPipe [alias-i.com]. A natural language processing program. Wonder if the two could be hooked up in kde or something hehe.

maxent.sf.net

[tealwarrior]

There is a mature statistical machine learning package on sourceforge. Check out maxent.sourceforge.net [sourceforge.net]. It's primarily been applied to natural language processing but it's applicable to a wide range of classification problems. There are even examples in the download package. I use it regularly and like it a lot but I'm also the primary maintainer so I might be biased.

Re:maxent.sf.net

[RandyF]

How well could maxent be applied to robotic independance? ie: training it to do a specific, but randomly encountered, task, like cleaning a room or picking up all the loose basketballs in the gym....

Re:maxent.sf.net

[tealwarrior]

The maxent code in this package classifies events with a descrete set of outcomes. You would have to break the above tasks down into descrete tasks. There might be very specific task like given features computed over some visual input, what angle (where 360 is divided into sufficiently small but descrete parts) should I approach the basketball for collection. Or if you have a routine to do one of those tasks you might use maxent to decide when you should enter that routine. It could compute isDirty for

Bah

[Anonymous Coward]

No support vector machines (indeed, any kind of kernel machines), POMDPs, AQ, NNs, Q-Learning, or stochastic optimization (EC, simulated annealing). Vunderbar.

my AI project

[OwlofCreamCheese]

takeing a intro to AI course this semester... made a program that can summerize encyclopdidia articals... sort of. all I know now is unicorns are made out of unicorn.

Weka

[auntfloyd]

There are already several excellent open source machine learning toolkits available. The one I have the most experience with is Weka [waikato.ac.nz], a Java-based system. In addition to providing an API, it has both command line and GUI tools.

With that and a decent ML book, I imagine most programmers could get up to speed rather quickly.

More info at DeviceForge

[Anonymous Coward]

Here's another article on DeviceForge [deviceforge.com] which includes an architecture graphic, a list of sub-projects (and links to them), links to a presentation about Intel's Open Source Machine Learning software and a technical whitepaper, and a link to the SourceForge download site.