On Wed, 13 Jun 2007 17:57:24 -0700, Richie <codesamur...@gmail.com> wrote: > I am a java programmer and I work in a rapid application development > environment at work. I am new to LISP and I am lost (because i have > been using limelight mainstream languages like java). I want to learn > and use LISP to develop AIs. Maybe, start with a simple game AI and > gradually advancing towards scientific or medical AIs and then to > financial AIs. There are several implementations or flavors of LISP. > Where do I start? Which ones used the most? or Does each > implementation have a specialty?
If you are used to Java, you might be familiar with the Eclipse IDE, and thus you might feel comfortabel starting with the cusp plugin:
I've bought the commercial version of LispWorks for Windows, and with Edi Weitz Starter Pack (http://weitz.de/starter-pack/) it's really wonderful, I'm still getting used to the EMACS bindings, so I'm also using OpenMCL/AquaMacs on my mac iBook, and also the Personal Edition of Lispworks there (which can't load the Edi Weitz Starter Pack automatically).
Richie wrote: > I am a java programmer and I work in a rapid application development > environment at work. I am new to LISP and I am lost (because i have > been using limelight mainstream languages like java). I want to learn > and use LISP to develop AIs. Maybe, start with a simple game AI and > gradually advancing towards scientific or medical AIs and then to > financial AIs. There are several implementations or flavors of LISP. > Where do I start? Which ones used the most? or Does each > implementation have a specialty?
Richie wrote: > I am a java programmer and I work in a rapid application development > environment at work. I am new to LISP and I am lost (because i have > been using limelight mainstream languages like java). I want to learn > and use LISP to develop AIs. Maybe, start with a simple game AI and > gradually advancing towards scientific or medical AIs and then to > financial AIs. There are several implementations or flavors of LISP. > Where do I start? Which ones used the most? or Does each > implementation have a specialty?
Welcome! :-D
Don't let the meanies (they'll prolly be here soon, if they haven't come already) scare you off -- they mean well. Just imagine them as troll repellant ;-)
First of all, LISP hasn't been spelled in all-caps for quite a while ;-) Just write Lisp. Second, Lisp is good for a lot of things besides AI. You can even develop web applications with it! Mostly I use it as an extension language for my C applications.
If you prefer Common Lisp, the student/trial versions of LispWorks or Allegro CL are quite easy to install. LispWorks has an IDE which will probably be familiar if you're used to point-and-click stuff. The open-source Lisps (of which there are many -- SBCL, CMUCL, OpenMCL, Clisp, ECL for starters) are typically used in the SLIME development environment under Emacs. SLIME is great -- it's well worth the initial learning curve.
If you prefer Scheme, the comp.lang.scheme newsgroup can help you :-) MIT and UC Berkeley, among other schools, teach intro programming using Scheme via the "Wizard Book" a.k.a. SICP (Structure and Interpretation of Computer Programs). There is a student-oriented Scheme IDE called DrScheme which is used for teaching Scheme.
I like Common Lisp macros better than Scheme macros, and I use macros a lot (I think most CL coders do). This is a plus for CL in my book. However, Scheme is a much more minimalist language. It may be useful if you are interested in embedded applications.
Richie wrote: > I am a java programmer and I work in a rapid application development > environment at work. I am new to LISP and I am lost (because i have > been using limelight mainstream languages like java). I want to learn > and use LISP to develop AIs. Maybe, start with a simple game AI and > gradually advancing towards scientific or medical AIs and then to > financial AIs. There are several implementations or flavors of LISP. > Where do I start? Which ones used the most? or Does each > implementation have a specialty?
Welcome! :-D
Don't let the meanies (they'll prolly be here soon, if they haven't come already) scare you off -- they mean well. Just imagine them as troll repellant ;-)
First of all, LISP hasn't been spelled in all-caps for quite a while ;-) Just write Lisp. Second, Lisp is good for a lot of things besides AI. You can even develop web applications with it! Mostly I use it as an extension language for my C applications.
If you prefer Common Lisp, the student/trial versions of LispWorks or Allegro CL are quite easy to install. LispWorks has an IDE which will probably be familiar if you're used to point-and-click stuff. The open-source Lisps (of which there are many -- SBCL, CMUCL, OpenMCL, Clisp, ECL for starters) are typically used in the SLIME development environment under Emacs. SLIME is great -- it's well worth the initial learning curve.
If you prefer Scheme, the comp.lang.scheme newsgroup can help you :-) MIT and UC Berkeley, among other schools, teach intro programming using Scheme via the "Wizard Book" a.k.a. SICP (Structure and Interpretation of Computer Programs). There is a student-oriented Scheme IDE called DrScheme which is used for teaching Scheme.
I like Common Lisp macros better than Scheme macros, and I use macros a lot (I think most CL coders do). This is a plus for CL in my book. However, Scheme is a much more minimalist language. It may be useful if you are interested in embedded applications.
Richie <codesamur...@gmail.com> writes: > I am a java programmer and I work in a rapid application development > environment at work. I am new to LISP and I am lost (because i have > been using limelight mainstream languages like java). I want to learn > and use LISP to develop AIs. Maybe, start with a simple game AI and > gradually advancing towards scientific or medical AIs and then to > financial AIs. There are several implementations or flavors of LISP. > Where do I start? Which ones used the most? or Does each > implementation have a specialty?
NOTE: The most fundamental particles in this product are held together by a "gluing" force about which little is currently known and whose adhesive power can therefore not be permanently guaranteed.
Richie wrote: > I am a java programmer and I work in a rapid application development > environment at work. I am new to LISP and I am lost (because i have > been using limelight mainstream languages like java). I want to learn > and use LISP to develop AIs. Maybe, start with a simple game AI and > gradually advancing towards scientific or medical AIs and then to > financial AIs. There are several implementations or flavors of LISP. > Where do I start? Which ones used the most? or Does each > implementation have a specialty?
Paradigms of Artificial Intelligence Programming by Peter Norvig is a great book for learning both classic AI techniques and Common Lisp at the same time.
On Jun 14, 2:57 am, Richie <codesamur...@gmail.com> wrote:
> I am a java programmer and I work in a rapid application development > environment at work. I am new to LISP and I am lost (because i have > been using limelight mainstream languages like java). I want to learn > and use LISP to develop AIs. Maybe, start with a simple game AI and > gradually advancing towards scientific or medical AIs and then to > financial AIs. There are several implementations or flavors of LISP. > Where do I start? Which ones used the most? or Does each > implementation have a specialty?
We don't do any AI since the AI winter in whenever that was. We use lisp only for web applications, accounting software and other real staff.
>There are several implementations or flavors of LISP. > Where do I start?
Get yourself a good book about lisp, then learn ai. Just search google group for tip about either and other posters already give you idea.
> Which ones used the most? or Does each > implementation have a specialty?
If you stick to the standard common lisp which you probably do since you're beginner just choose whichever fits for you .
Some vendors (in alphabetical order) Corman (only windows) Franz Lispworks Scieneer (only linux)
Some free implementations: Clisp CMUCL ECL OpenMCL SBCL
If you tell us what platform you prefer win,linux etc you'll probably get a more specific advice.
On 14 Jun, 08:55, fireblade <slobodan.blaze...@gmail.com> wrote:
> We don't do any AI since the AI winter in whenever that was. We use > lisp only for web applications, accounting software and other real > staff.
That's just what we want you to think till the robots take over.
Seriously, AI didn't go away, it took a serious kicking, rid itself of some of it's old baggage (some good,some bad)* and re-branded itself as machine learning.
*Disclaimer I wasn't around for the AI winter, so I don't know quite how much cool stuff was thrown out then.
> On Wed, 13 Jun 2007 17:57:24 -0700, Richie <codesamur...@gmail.com> wrote: > > I am a java programmer and I work in a rapid application development > > environment at work. I am new to LISP and I am lost (because i have > > been using limelight mainstream languages like java). I want to learn > > and use LISP to develop AIs. Maybe, start with a simple game AI and > > gradually advancing towards scientific or medical AIs and then to > > financial AIs. There are several implementations or flavors of LISP. > > Where do I start? Which ones used the most? or Does each > > implementation have a specialty?
> If you are used to Java, you might be familiar with the Eclipse IDE, and > thus you might feel comfortabel starting with the cusp plugin:
I would also recommend Peter Seidel's Lisp book and would also recommend Paul Graham's Ansi Common Lisp and once your more advanced I would recommend OnLisp Awsome the book is now a cult his and routinely sells for $200+ but Paul is giving away an online version for free.
I find lisp is one of those languages where you have to read a few books to get the hang of it.
> On 14 Jun., 03:21, "Peter Hildebrandt" <pet...@icsi.berkeley.edu> > wrote:
> > On Wed, 13 Jun 2007 17:57:24 -0700, Richie <codesamur...@gmail.com> wrote: > > > I am a java programmer and I work in a rapid application development > > > environment at work. I am new to LISP and I am lost (because i have > > > been using limelight mainstream languages like java). I want to learn > > > and use LISP to develop AIs. Maybe, start with a simple game AI and > > > gradually advancing towards scientific or medical AIs and then to > > > financial AIs. There are several implementations or flavors of LISP. > > > Where do I start? Which ones used the most? or Does each > > > implementation have a specialty?
> > If you are used to Java, you might be familiar with the Eclipse IDE, and > > thus you might feel comfortabel starting with the cusp plugin:
On Jun 14, 2:22 pm, Chris Russell <christopher.m.russ...@gmail.com> wrote:
> On 14 Jun, 08:55, fireblade <slobodan.blaze...@gmail.com> wrote:
> > We don't do any AI since the AI winter in whenever that was. We use > > lisp only for web applications, accounting software and other real > > staff.
> That's just what we want you to think till the robots take over.
And that will be ? Insert year____
> Seriously, AI didn't go away, it took a serious kicking, rid itself of > some of it's old baggage (some good,some bad)* and re-branded itself > as machine learning.
Data mining perhaps. It's all written in Java.
> *Disclaimer I wasn't around for the AI winter, so I don't know quite > how much cool stuff was thrown out then.
Yesterday I've read Richard P. Gabriels Patterns of software so got some idea about those *fabolous* times. Actually till yesterday I didn't know that he started Lucid.All I've knew about him is the essey "Worse is better."
On Jun 14, 1:04 pm, Richie <codesamur...@gmail.com> wrote:
> i am using windows XP Pro platform.
> Thanks all for the information. if you have more information, please > keep it coming.
If you like slower pace try gentle introduction to symbolic computation. Printed is hard to find but there's free pdf at authors site . Fast going Paul Graham Ansi Common Lisp. If you just want to go directly to practical staff, Practical Common Lisp from Seibel. On Lisp is excellent tutorial for macros. I can't recommend anything for AI , probably Norvigs PAIP & AI modern approach (wuth Rassel) , Kanemoto also looked interesthing but I haven't try them.
On Fri, 15 Jun 2007 00:35:43 -0700, fireblade <slobodan.blaze...@gmail.com> wrote: > Does hunchentoot runs on scieneer ?
I'd be surprised if it did, because there's no support code for Scieneer CL in Hunchentoot's code base yet (patches welcome, of course) and I don't have a copy of Scieneer CL to port it myself.
Also, Scieneer CL has integrated HTTP support, so it would probably be worthwhile to let Scieneer CL do some of the work Hunchentoot does on other Lisps.
Anyway, if you're interested, you should contact them. They have ported a lot of open source Lisp libraries to their Lisp (in fact, Douglas Crosher just sent patches for CXML the other day) and I'm sure they'd be happy to port others as well if there's customer demand.
--
Lisp is not dead, it just smells funny.
Real email: (replace (subseq "spamt...@agharta.de" 5) "edi")
> On Fri, 15 Jun 2007 00:35:43 -0700, fireblade <slobodan.blaze...@gmail.com> wrote: > > Does hunchentoot runs on scieneer ?
> I'd be surprised if it did, because there's no support code for > Scieneer CL in Hunchentoot's code base yet (patches welcome, of > course) and I don't have a copy of Scieneer CL to port it myself.
> Also, Scieneer CL has integrated HTTP support, so it would probably be > worthwhile to let Scieneer CL do some of the work Hunchentoot does on > other Lisps.
> Anyway, if you're interested, you should contact them. They have > ported a lot of open source Lisp libraries to their Lisp (in fact, > Douglas Crosher just sent patches for CXML the other day) and I'm sure > they'd be happy to port others as well if there's customer demand.
> --
> Lisp is not dead, it just smells funny.
> Real email: (replace (subseq "spamt...@agharta.de" 5) "edi")
Currently I'm developing on windows only (ok sometimes I try it on Ubuntu too), but I would be happier with a linux box if there will be someone to administer it, currently our servers are windows only. Scieneer is in budget also and supports SMP so it would be interesthing how it compares with lw.
On 15 Jun, 08:41, fireblade <slobodan.blaze...@gmail.com> wrote:
> On Jun 14, 2:22 pm, Chris Russell <christopher.m.russ...@gmail.com> > wrote:> On 14 Jun, 08:55, fireblade <slobodan.blaze...@gmail.com> wrote:
> > > We don't do any AI since the AI winter in whenever that was. We use > > > lisp only for web applications, accounting software and other real > > > staff.
> > That's just what we want you to think till the robots take over.
> And that will be ? Insert year____
Oh, you'll find out.....
> > Seriously, AI didn't go away, it took a serious kicking, rid itself of > > some of it's old baggage (some good,some bad)* and re-branded itself > > as machine learning.
> Data mining perhaps. It's all written in Java.
Meh, I'm not sure I'd count data mining as machine learning, but I suppose the more sophisticated of it must be. The computer vision used in scene understanding, generating 3-d models from still images or self-parking cars might be better examples.
> On 15 Jun, 08:41, fireblade <slobodan.blaze...@gmail.com> wrote:> On Jun 14, 2:22 pm, Chris Russell <christopher.m.russ...@gmail.com> > > wrote:> On 14 Jun, 08:55, fireblade <slobodan.blaze...@gmail.com> wrote:
> > > > We don't do any AI since the AI winter in whenever that was. We use > > > > lisp only for web applications, accounting software and other real > > > > staff.
> > > That's just what we want you to think till the robots take over.
> > And that will be ? Insert year____
> Oh, you'll find out..... Hope so
> > > Seriously, AI didn't go away, it took a serious kicking, rid itself of > > > some of it's old baggage (some good,some bad)* and re-branded itself > > > as machine learning.
> > Data mining perhaps. It's all written in Java.
> Meh, I'm not sure I'd count data mining as machine learning, but I > suppose the more sophisticated of it must be.
Clustering,rules generarting expert systems, pattern matching, fuzzy- logic, genetic algorithms .. anything of those sounds familiar ? All of those comes even in a free for learning purposes application like WEKA.
> The computer vision used in scene understanding, generating 3-d models > from still images or self-parking cars might be better examples.
I'm too lazy to search but I'll bet again on Java for all of above.
> > > > Seriously, AI didn't go away, it took a serious kicking, rid itself of > > > > some of it's old baggage (some good,some bad)* and re-branded itself > > > > as machine learning.
> > > Data mining perhaps. It's all written in Java.
> > Meh, I'm not sure I'd count data mining as machine learning, but I > > suppose the more sophisticated of it must be.
> Clustering,rules generarting expert systems, pattern matching, fuzzy- > logic, genetic algorithms .. anything of those sounds familiar ?
Yes, but its somewhat disorientating. I've no idea why you're classifying pattern matching and the vast majority of clustering as AI, and I assumed that fuzzy logic and expert systems were dead and gone.
> > The computer vision used in scene understanding, generating 3-d models > > from still images or self-parking cars might be better examples.
> I'm too lazy to search but I'll bet again on Java for all of above.
Largely matlab and c[++] when people care about speed or hate matlab.
On Jun 18, 12:05 pm, Chris Russell <christopher.m.russ...@gmail.com> wrote:
> > > > > Seriously, AI didn't go away, it took a serious kicking, rid itself of > > > > > some of it's old baggage (some good,some bad)* and re-branded itself > > > > > as machine learning.
> > > > Data mining perhaps. It's all written in Java.
> > > Meh, I'm not sure I'd count data mining as machine learning, but I > > > suppose the more sophisticated of it must be.
On Jun 18, 3:24 pm, Chris Russell <christopher.m.russ...@gmail.com> wrote:
> On 18 Jun, 12:46, fireblade <slobodan.blaze...@gmail.com> wrote:
> > On Jun 18, 12:05 pm, Chris Russell <christopher.m.russ...@gmail.com> > > >and I assumed that fuzzy logic and expert systems were dead and gone.
> > Obviously not, there's big money involved.
> Cool, what are they being used for?
Discovery of patterns hidden in large databases(warehouses). Did I mentioned neural networks? No ? Those too. Here's the way it goes: 1. (Naive) Bayesian first 2. Clustering second 3. Fuzzy logic third 4. Neural networks fourth.(hardest patterns to use)
There is also genetic algorithms used but they are more of a hm search techniques.
Chris Russell <christopher.m.russ...@gmail.com> writes: > On 18 Jun, 12:46, fireblade <slobodan.blaze...@gmail.com> wrote: >> On Jun 18, 12:05 pm, Chris Russell <christopher.m.russ...@gmail.com>
>> >and I assumed that fuzzy logic and expert systems were dead and gone.
>> Obviously not, there's big money involved.
> Cool, what are they being used for?
As for expert systems, my understanding is that they've gone mainstream in some areas, such as tax planning, which seems like a natural match. (I think I learned this tidbit a few years ago during a conversation with Ed Feigenbaum, who's in a position to know this sort of thing.)
