Hello, It is so good to hear that there is an external for multi-layered neural networks. I am doing a research about simulating human musical interaction where I am building a machine that can learn and perform music in real time.
Davide, it would be nice if you can announce to the list when you have makefile or pre-compiled external of nn for OSX.
and Davide and IOhannes, do you have any written documentations for nn and ann_som externals, something like more than help files?
best, Koray.
On May 9, 2005, at 1:30, pd-list-request@iem.at wrote:
Message: 8 Date: Sun, 8 May 2005 19:20:41 +0200 (MEST) From: "Georg Holzmann" grhPD@gmx.at Subject: Re: [PD] New external on neural networks To: info@davidemorelli.it Cc: pd-list@iem.at Message-ID: 3210.1115572841@www37.gmx.net Content-Type: text/plain; charset="us-ascii"
Hallo!
I just finished a little external I wrote to use FANN classes in PD. It is about multi layered neural networks.
cool - I also wanted to do that ... :)
I would like to know who is the author of the ann_som external, I think he could be interested in it (for ann_mlp).
it's iohannes.
It should compile on linux and osx also but I didn't try (you'll need to
add
paths to fann src and libs).. if you do it please send me a makefile and I'll add it.
the makefile for linux is at grh.mur.at/Makefile, a minor change to compile nn.c on linux at grh.mur.at/nn.c
Maybe you can put this in the externals/ann folder in cvs?
LG Georg
M.Koray Tahiroglu Media Lab, University of Art and Design Helsinki, TaiK Hameentie 135C 00560 Helsinki Finland http://mlab.uiah.fi/~korayt/ http://purenoise.uiah.fi:8000/ tel: +358 40 754 8449 fax: +358 9 75630 555
hello, I think there is a survey about that topic at the homepage of la-kitchen (Arshia Cont, Thierry Coduys, Cyrille Henry, ...) http://www.la-kitchen.fr/augmented_mapping.html. and there are some links for literature upon that topic. they also put some patches on their website. I stumbled about the mtx objects, which are no more in zexy. but that will be a seperate posting. marius.
----- Original Message ----- From: "Koray Tahiroglu" ktahirog@uiah.fi To: pd-list@iem.at Cc: info@davidemorelli.it; "IOhannes m zmoelnig" zmoelnig@iem.at Sent: Monday, May 09, 2005 10:03 AM Subject: Re: [PD] New external on neural networks
Hello, It is so good to hear that there is an external for multi-layered neural networks. I am doing a research about simulating human musical interaction where I am building a machine that can learn and perform music in real time.
Davide, it would be nice if you can announce to the list when you have makefile or pre-compiled external of nn for OSX.
and Davide and IOhannes, do you have any written documentations for nn and ann_som externals, something like more than help files?
best, Koray.
On May 9, 2005, at 1:30, pd-list-request@iem.at wrote:
Message: 8 Date: Sun, 8 May 2005 19:20:41 +0200 (MEST) From: "Georg Holzmann" grhPD@gmx.at Subject: Re: [PD] New external on neural networks To: info@davidemorelli.it Cc: pd-list@iem.at Message-ID: 3210.1115572841@www37.gmx.net Content-Type: text/plain; charset="us-ascii"
Hallo!
I just finished a little external I wrote to use FANN classes in PD. It is about multi layered neural networks.
cool - I also wanted to do that ... :)
I would like to know who is the author of the ann_som external, I think he could be interested in it (for ann_mlp).
it's iohannes.
It should compile on linux and osx also but I didn't try (you'll need to
add
paths to fann src and libs).. if you do it please send me a makefile and I'll add it.
the makefile for linux is at grh.mur.at/Makefile, a minor change to compile nn.c on linux at grh.mur.at/nn.c
Maybe you can put this in the externals/ann folder in cvs?
LG Georg
M.Koray Tahiroglu Media Lab, University of Art and Design Helsinki, TaiK Hameentie 135C 00560 Helsinki Finland http://mlab.uiah.fi/~korayt/ http://purenoise.uiah.fi:8000/ tel: +358 40 754 8449 fax: +358 9 75630 555
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
hello,
Marius Schebella a écrit :
hello, I think there is a survey about that topic at the homepage of la-kitchen (Arshia Cont, Thierry Coduys, Cyrille Henry, ...) http://www.la-kitchen.fr/augmented_mapping.html. and there are some links for literature upon that topic.
Arshia patch's does only use a matrix as neuronal coeficient. The training of the network was made with matlab. So it's not a only pd solution. nothing really finish here, and there is no more development on this project.
they also put some patches on their website. I stumbled about the mtx objects, which are no more in zexy.
mtx are really usefull!
cyrille
but that will be a seperate posting. marius.
----- Original Message ----- From: "Koray Tahiroglu" ktahirog@uiah.fi To: pd-list@iem.at Cc: info@davidemorelli.it; "IOhannes m zmoelnig" zmoelnig@iem.at Sent: Monday, May 09, 2005 10:03 AM Subject: Re: [PD] New external on neural networks
Hello, It is so good to hear that there is an external for multi-layered neural networks. I am doing a research about simulating human musical interaction where I am building a machine that can learn and perform music in real time.
Davide, it would be nice if you can announce to the list when you have makefile or pre-compiled external of nn for OSX.
and Davide and IOhannes, do you have any written documentations for nn and ann_som externals, something like more than help files?
best, Koray.
On May 9, 2005, at 1:30, pd-list-request@iem.at wrote:
Message: 8 Date: Sun, 8 May 2005 19:20:41 +0200 (MEST) From: "Georg Holzmann" grhPD@gmx.at Subject: Re: [PD] New external on neural networks To: info@davidemorelli.it Cc: pd-list@iem.at Message-ID: 3210.1115572841@www37.gmx.net Content-Type: text/plain; charset="us-ascii"
Hallo!
I just finished a little external I wrote to use FANN classes in PD. It is about multi layered neural networks.
cool - I also wanted to do that ... :)
I would like to know who is the author of the ann_som external, I think he could be interested in it (for ann_mlp).
it's iohannes.
It should compile on linux and osx also but I didn't try (you'll need to
add
paths to fann src and libs).. if you do it please send me a makefile and I'll add it.
the makefile for linux is at grh.mur.at/Makefile, a minor change to compile nn.c on linux at grh.mur.at/nn.c
Maybe you can put this in the externals/ann folder in cvs?
LG Georg
M.Koray Tahiroglu Media Lab, University of Art and Design Helsinki, TaiK Hameentie 135C 00560 Helsinki Finland http://mlab.uiah.fi/~korayt/ http://purenoise.uiah.fi:8000/ tel: +358 40 754 8449 fax: +358 9 75630 555
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Hallo, Koray Tahiroglu hat gesagt: // Koray Tahiroglu wrote:
and Davide and IOhannes, do you have any written documentations for nn and ann_som externals, something like more than help files?
The FANN library, Davide uses, has huge amounts of documentation on http://fann.sourceforge.net/
Ciao
Frank Barknecht _ ______footils.org__
_ __latest track: "scans" _ http://footils.org/cms/show/41
IOhannes: the learn_from_file() function must be called only once, after the net is trained you can save its state sending 2 messages: [filename test.net( then [save( the net can be loaded simply using the test.net as argument [nn test.net] or by sending 2 messages: [filename test.net( and [load(
IOhannes and Georg: I would really like to have on-line training but I think is not supported by FANN right now.. umph.
I never created an autoconf myself although I saw many, I will try as soon as I have some spare time. (I must find the time to reinstall linux at home) It will be a perfect solution for linux and osx. ...but what about win32?
Koray: about resources on Neural Networks have a look at this wiki page I am trying to feed with every link I find interesting on AI: http://puredata.org/Members/dmorelli/AIResources
----- Original Message ----- From: "Koray Tahiroglu" ktahirog@uiah.fi To: pd-list@iem.at Cc: info@davidemorelli.it; "IOhannes m zmoelnig" zmoelnig@iem.at Sent: Monday, May 09, 2005 10:03 AM Subject: Re: [PD] New external on neural networks
Hello, It is so good to hear that there is an external for multi-layered neural networks. I am doing a research about simulating human musical interaction where I am building a machine that can learn and perform music in real time.
Davide, it would be nice if you can announce to the list when you have makefile or pre-compiled external of nn for OSX.
and Davide and IOhannes, do you have any written documentations for nn and ann_som externals, something like more than help files?
best, Koray.
On May 9, 2005, at 1:30, pd-list-request@iem.at wrote:
Message: 8 Date: Sun, 8 May 2005 19:20:41 +0200 (MEST) From: "Georg Holzmann" grhPD@gmx.at Subject: Re: [PD] New external on neural networks To: info@davidemorelli.it Cc: pd-list@iem.at Message-ID: 3210.1115572841@www37.gmx.net Content-Type: text/plain; charset="us-ascii"
Hallo!
I just finished a little external I wrote to use FANN classes in PD. It is about multi layered neural networks.
cool - I also wanted to do that ... :)
I would like to know who is the author of the ann_som external, I think he could be interested in it (for ann_mlp).
it's iohannes.
It should compile on linux and osx also but I didn't try (you'll need to
add
paths to fann src and libs).. if you do it please send me a makefile and I'll add it.
the makefile for linux is at grh.mur.at/Makefile, a minor change to compile nn.c on linux at grh.mur.at/nn.c
Maybe you can put this in the externals/ann folder in cvs?
LG Georg
M.Koray Tahiroglu Media Lab, University of Art and Design Helsinki, TaiK Hameentie 135C 00560 Helsinki Finland http://mlab.uiah.fi/~korayt/ http://purenoise.uiah.fi:8000/ tel: +358 40 754 8449 fax: +358 9 75630 555
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Davide Morelli wrote:
IOhannes: the learn_from_file() function must be called only once, after the net is trained you can save its state sending 2 messages: [filename test.net( then [save( the net can be loaded simply using the test.net as argument [nn test.net] or by sending 2 messages: [filename test.net( and [load(
yes, i have seen that as soon as i have compiled and tried your external. (btw, wouldn't it be nice to also accept [save test.net( instead of [filename test.net, save( ??
IOhannes and Georg: I would really like to have on-line training but I think is not supported by FANN right now.. umph.
looking at my fann.h file (from fann-1.2.0) i see a fann_train()-function which should do that, no ? (this will work only for floating-point values though)
I never created an autoconf myself although I saw many, I will try as soon as I have some spare time. (I must find the time to reinstall linux at home) It will be a perfect solution for linux and osx. ...but what about win32?
cygwin will know how to handle it. a plain msvc-environment is different of course, but i guess one could live with that:
just an idea:
- provide a "Makefile" for msvc (which people would have to edit to
match their system)
- provide a configure script that overwrites the "Makefile" on systems
that support configure
or: use "Makefile.nt"
this reminds me, that i should buy a written documentation (aka: handbook) for make.
mfg.a.dr IOhannes
[save filename( [load filename( Ok, done.
fann_train() I am trying to implement on-the-fly learning but the net seems to reset each time I train it.. Working on it..
autoconf: I agree, I'll try to write
- an autoconf
AND
- a Makefile for msvc (the one I'm using now)
Btw: how does one get access to CVS as a developer?
-----Messaggio originale----- Da: pd-list-bounces@iem.at [mailto:pd-list-bounces@iem.at] Per conto di IOhannes m zmoelnig Inviato: lunedì 9 maggio 2005 14.18 A: Davide Morelli Cc: PD Lista Oggetto: Re: [PD] New external on neural networks
Davide Morelli wrote:
IOhannes: the learn_from_file() function must be called only once, after the net is trained you can save its state sending 2 messages: [filename test.net( then [save( the net can be loaded simply using the test.net as argument [nn test.net] or by sending 2 messages: [filename test.net( and [load(
yes, i have seen that as soon as i have compiled and tried your external. (btw, wouldn't it be nice to also accept [save test.net( instead of [filename test.net, save( ??
IOhannes and Georg: I would really like to have on-line training but I think is not supported by FANN right now.. umph.
looking at my fann.h file (from fann-1.2.0) i see a fann_train()-function which should do that, no ? (this will work only for floating-point values though)
I never created an autoconf myself although I saw many, I
will try as
soon as I have some spare time. (I must find the time to
reinstall linux
at home) It will be a perfect solution for linux and osx. ...but
what about win32?
cygwin will know how to handle it. a plain msvc-environment is different of course, but i guess one could live with that:
just an idea:
- provide a "Makefile" for msvc (which people would have to edit to
match their system)
- provide a configure script that overwrites the "Makefile"
on systems that support configure
or: use "Makefile.nt"
this reminds me, that i should buy a written documentation (aka: handbook) for make.
mfg.a.dr IOhannes
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-- No virus found in this incoming message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.3 - Release Date: 03/05/2005
Hi list,
I just compiled a new version of nn
I included: -on the fly training -save and load the ann in 1 click
Can be downloaded from here: https://www.puredata.info/Members/dmorelli/nn%20v-0.02/download
The next version will be renamed from "nn" to "ann_mlp", will be part of the ann "library" and will be put in cvs in the ann folder.
-----Messaggio originale----- Da: pd-list-bounces@iem.at [mailto:pd-list-bounces@iem.at] Per conto di IOhannes m zmoelnig Inviato: lunedì 9 maggio 2005 14.18 A: Davide Morelli Cc: PD Lista Oggetto: Re: [PD] New external on neural networks
Davide Morelli wrote:
IOhannes: the learn_from_file() function must be called only once, after the net is trained you can save its state sending 2 messages: [filename test.net( then [save( the net can be loaded simply using the test.net as argument [nn test.net] or by sending 2 messages: [filename test.net( and [load(
yes, i have seen that as soon as i have compiled and tried your external. (btw, wouldn't it be nice to also accept [save test.net( instead of [filename test.net, save( ??
IOhannes and Georg: I would really like to have on-line training but I think is not supported by FANN right now.. umph.
looking at my fann.h file (from fann-1.2.0) i see a fann_train()-function which should do that, no ? (this will work only for floating-point values though)
I never created an autoconf myself although I saw many, I
will try as
soon as I have some spare time. (I must find the time to
reinstall linux
at home) It will be a perfect solution for linux and osx. ...but
what about win32?
cygwin will know how to handle it. a plain msvc-environment is different of course, but i guess one could live with that:
just an idea:
- provide a "Makefile" for msvc (which people would have to edit to
match their system)
- provide a configure script that overwrites the "Makefile"
on systems that support configure
or: use "Makefile.nt"
this reminds me, that i should buy a written documentation (aka: handbook) for make.
mfg.a.dr IOhannes
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-- No virus found in this incoming message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.3 - Release Date: 03/05/2005
Hallo, Davide Morelli hat gesagt: // Davide Morelli wrote:
The next version will be renamed from "nn" to "ann_mlp",
Just wondering: Why not call it "fann"? Less typing is a Good Thing and AFAIK "fann" is still free.
Ciao
Frank Barknecht _ ______footils.org__
_ __latest track: "scans" _ http://footils.org/cms/show/41
That was my first idea, then I chose nn because "Less typing is a Good Thing". Now I think ann_mlp is a good idea because: ann_som is about self organized maps ann_mlp is about multilayer perceptrons (waiting for ann_rfb etc..) This way all names are unified and coders know that ann_ is where they should look for Neural Networks.
I personally like both "fann" and "ann_mlp"..
Let's start a poll!
Do you prefer brevity or uniformity?
-----Messaggio originale----- Da: pd-list-bounces@iem.at [mailto:pd-list-bounces@iem.at] Per conto di Frank Barknecht Inviato: lunedì 9 maggio 2005 23.02 A: pd-list@iem.at Oggetto: Re: R: [PD] New external on neural networks
Hallo, Davide Morelli hat gesagt: // Davide Morelli wrote:
The next version will be renamed from "nn" to "ann_mlp",
Just wondering: Why not call it "fann"? Less typing is a Good Thing and AFAIK "fann" is still free.
Ciao
Frank Barknecht _ ______footils.org__
_ __latest track: "scans" _ http://footils.org/cms/show/41
PD-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
-- No virus found in this incoming message. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.3 - Release Date: 03/05/2005
Hallo, Davide Morelli hat gesagt: // Davide Morelli wrote:
That was my first idea, then I chose nn because "Less typing is a Good Thing". Now I think ann_mlp is a good idea because: ann_som is about self organized maps ann_mlp is about multilayer perceptrons (waiting for ann_rfb etc..) This way all names are unified and coders know that ann_ is where they should look for Neural Networks.
Okay, that makes sense. I admit it's quite some time I looked at "ann". From a typing perspective however underscores are a pain at least on german keyboards as you need to hold "Shift", so I would prefer the dot as in "ann.som" or "vasp.something" but that's a minor complain.
Ciao
Frank Barknecht _ ______footils.org__
_ __latest track: "scans" _ http://footils.org/cms/show/41
-
'Frank Barknecht' -
cyrille henry -
Davide Morelli -
Frank Barknecht
-
IOhannes m zmoelnig -
Koray Tahiroglu -
Marius Schebella