Visualisation of Machine Learning Algorithms

#1

Post by maurits » Mon, May 30, 2011, 12:53 PM UTC

Visualisation of Machine Learning Algorithms
mldemos.epfl.ch

Re: Visualisation of Machine Learning Algorithms

#2

Post by vinyl » Mon, May 30, 2011, 1:37 PM UTC

Wow. This could prove an invaluable tool for teaching. Great UI (love the way you can build the dataset with a paintbrush), top-quality visualization, wide choice of algorithms... Much better than R scripts if you want to show basic algorithms at work to a bunch of students. Many thanks

Re: Visualisation of Machine Learning Algorithms

#3

Post by maeon3 » Mon, May 30, 2011, 2:18 PM UTC

Patch through a black and white video into this tool and watch how the algorithm tries to find patterns in a 2d reprsentation of a 3d space. Maybe it can predict where the table ends without full view of the table.

Neat stuff, it is the beginning of machine notion of object permanence. Filling in the gaps and making educated guesses about what kind of function controls where a sequence originated from.

Re: Visualisation of Machine Learning Algorithms

#4

Post by vedantk » Mon, May 30, 2011, 2:40 PM UTC

The install.sh and install_script.sh scripts are Mac-dependent. How would you compile this on Linux? Qmake?

edit: Yep. Clone the git repo, then run qmake, followed by make.

edit #2: Here is a patch I wrote to make the project compile under OpenCV 2.2, and Linux: https://gist.github.com/999061

Re: Visualisation of Machine Learning Algorithms

#5

Post by vedantk » Mon, May 30, 2011, 2:58 PM UTC

vedantk wrote: post #4

The install.sh and install_script.sh scripts are Mac-dependent. How would you compile this on Linux? Qmake? edit: Yep. Clone the git repo, then run qmake, followed by make. edit #2: Here is a patch I wrote to make the project compile under OpenCV 2.2, and Linux: https://gist.github.com/999061

Here's a patch that lets you compile with opencv2: https://gist.github.com/999061.

edit: I'm still getting a Make error for cvSumPixels (unless I comment that check out);

sampleManager.cpp:(.text+0x1a06): undefined reference to `cvSumPixels'

Re: Visualisation of Machine Learning Algorithms

#6

Post by b4silio » Mon, May 30, 2011, 3:56 PM UTC

vedantk wrote: post #5

vedantk wrote: post #4

The install.sh and install_script.sh scripts are Mac-dependent. How would you compile this on Linux? Qmake? edit: Yep. Clone the git repo, then run qmake, followed by make. edit #2: Here is a patch I wrote to make the project compile under OpenCV 2.2, and Linux: https://gist.github.com/999061

Here's a patch that lets you compile with opencv2: https://gist.github.com/999061 . edit: I'm still getting a Make error for cvSumPixels (unless I comment that check out); sampleManager.cpp:(.text+0x1a06): undefined reference to `cvSumPixels'

you might need to add #include #include to get those I haven't gotten around to porting it to opencv2.2 so the patch is very welcome!

Re: Visualisation of Machine Learning Algorithms

#7

Post by b4silio » Mon, May 30, 2011, 4:05 PM UTC

vedantk wrote: post #4

The install.sh and install_script.sh scripts are Mac-dependent. How would you compile this on Linux? Qmake? edit: Yep. Clone the git repo, then run qmake, followed by make. edit #2: Here is a patch I wrote to make the project compile under OpenCV 2.2, and Linux: https://gist.github.com/999061

Yeah, the install(_script).sh are to allow deployment on the Mac (which wants absolute paths to libraries). They will need to be changed to direct to the local libraries in case you need them... (hopefully a new version of MacDeployQt will appear at some point to take care of all that)

Re: Visualisation of Machine Learning Algorithms

#8

Post by vedantk » Mon, May 30, 2011, 8:38 PM UTC

b4silio wrote: post #6

vedantk wrote: post #5

Earlier quoted context omitted.

Here's a patch that lets you compile with opencv2: https://gist.github.com/999061 . edit: I'm still getting a Make error for cvSumPixels (unless I comment that check out); sampleManager.cpp:(.text+0x1a06): undefined reference to `cvSumPixels'

you might need to add #include #include to get those I haven't gotten around to porting it to opencv2.2 so the patch is very welcome!

Good call. I added those headers and linked to their related libraries. It's working perfectly on my Arch install.

Here are my modifications; https://gist.github.com/999427

Re: Visualisation of Machine Learning Algorithms

#9

Post by Emore » Mon, May 30, 2011, 9:38 PM UTC

Here's another one, a simple Applet that lets you apply various classification algorithms on 2D data: http://www.cs.technion.ac.il/~rani/LocBoost/

Re: Visualisation of Machine Learning Algorithms

#10

Post by microsage » Mon, May 30, 2011, 10:37 PM UTC

vedantk wrote: post #4

The install.sh and install_script.sh scripts are Mac-dependent. How would you compile this on Linux? Qmake? edit: Yep. Clone the git repo, then run qmake, followed by make. edit #2: Here is a patch I wrote to make the project compile under OpenCV 2.2, and Linux: https://gist.github.com/999061

Even after applying your patch I had a little trouble getting everything to link properly on x86_64 Fedora (14). I ended up manually tweaking my make files after running qmake.

Changed the LIBS line (in MLDemos/Makefile, _AlgorithmPlugins//Makefile., _IOPlugins//Makefile.) to:

LIBS = $(SUBLIBS) -L/usr/lib64 -L/usr/local/lib -lhighgui -lcv -lcxcore -lQtGui -lQtCore -lpthread

I suspect there's an easier way, but this at least allowed me to build and everything seems to work properly.