Methodology
Locked
 
|
Mohamed,
I like the idea of a working group to start exploring an appropriate methodology for ABMs. I think it's time we tried to get some intellectual clout behind this whole question of validity. As to how this could proceed, CRISP-DM and KADS may provide some pointers, albeit they are both European programs. Both started from informal working groups involving (on the push end) software vendors, consultants or academics and (on the pull end) enlightened clients. Ultimately, both attracted European Community funding (through the ESPRIT program). In fact KADS attracted a huge pile of funds - the program has been running since 1990 and is still underway. Robert P.S. Just to give you a flavour of the groups' scope, here are their members: KADS-II (ESPRIT Project P5248) Cap Gemini Innovation Swedish Institute for Computer Science Cap Programator Siemens AG Netherlands Energy Research Foundation ECN Touche Ross Management Consultants Entel SA University of Amsterdam IBM France Free University Brussels Lloyd's Register CRISP_DM (ESPRIT Project 24.959) Teradata (Denmark) SPSS (was ISL) - vendors of 'Clementine' data-mining software DaimlerChrysler OHRA (Netherlands) |
|
|
I'm certainly interested in having ABM "grow up", so to speak. I've
been working with the Value Network working group at SFI: http://valuenetwork.sourceforge.net/ http://backspaces.net/SCSim/ and part of the work has been looking at validating models. One approach taken by the Argonne folks was to implement the model in multiple standard environments .. called "docking" http://backspaces.net/SCSim/BeerDock.pdf However, these are more low level apprroaches: how to be sure the algorithms are robust and repeatable. Deeper questions lurk: how to build models that reflect the situation being modeled accurately. I'd love to play! Owen On Tuesday, May 6, 2003, at 09:25 AM, Robert Holmes wrote: > Mohamed, > > I like the idea of a working group to start exploring an appropriate > methodology for ABMs. I think it's time we tried to get some > intellectual > clout behind this whole question of validity. > > As to how this could proceed, CRISP-DM and KADS may provide some > pointers, > albeit they are both European programs. Both started from informal > working > groups involving (on the push end) software vendors, consultants or > academics and (on the pull end) enlightened clients. Ultimately, both > attracted European Community funding (through the ESPRIT program). In > fact > KADS attracted a huge pile of funds - the program has been running > since > 1990 and is still underway. > > Robert > > P.S. Just to give you a flavour of the groups' scope, here are their > members: > > KADS-II (ESPRIT Project P5248) > Cap Gemini Innovation > Swedish Institute for Computer Science > Cap Programator > Siemens AG > Netherlands Energy Research Foundation ECN > Touche Ross Management Consultants > Entel SA > University of Amsterdam > IBM France > Free University Brussels > Lloyd's Register > > CRISP_DM (ESPRIT Project 24.959) > Teradata (Denmark) > SPSS (was ISL) - vendors of 'Clementine' data-mining software > DaimlerChrysler > OHRA (Netherlands) > > > > > > > ========================================================= > FRIAM Complexity Coffee listserv > Meets Fridays 9AM @ Museum Hill Cafe > Archives, unsubscribe, etc.: > http://www.redfish.com/mailman/listinfo/friam_redfish.com > Work: 505-983-6305 Cell: 505-570-0168 Home: 505-988-3787 [hidden email] http://backspaces.net http://complexityworkshop.com |
|
|
Robert and Owen,
Maybe a good first step might be that we and fellow FRIAMers should start putting some ideas together. Publish it and try to get industrial and governmental buy-in. How about we chat about it? I would be happy to host it http://www.optomatica.com/ABM-M.htm Lets agree on the time... how about next Monday 9 MST? I am about 9 hrs ahead of you, so the earlier the better. Mohammed ----- Original Message ----- From: "Owen Densmore" <[hidden email]> To: <[hidden email]> Sent: Tuesday, May 06, 2003 8:18 PM Subject: Re: [FRIAM] Methodology > I'm certainly interested in having ABM "grow up", so to speak. I've > been working with the Value Network working group at SFI: > http://valuenetwork.sourceforge.net/ > http://backspaces.net/SCSim/ > and part of the work has been looking at validating models. One > approach taken by the Argonne folks was to implement the model in > multiple standard environments .. called "docking" > http://backspaces.net/SCSim/BeerDock.pdf > > However, these are more low level apprroaches: how to be sure the > algorithms are robust and repeatable. Deeper questions lurk: how to > build models that reflect the situation being modeled accurately. > > I'd love to play! > > Owen > > On Tuesday, May 6, 2003, at 09:25 AM, Robert Holmes wrote: > > > Mohamed, > > > > I like the idea of a working group to start exploring an appropriate > > methodology for ABMs. I think it's time we tried to get some > > intellectual > > clout behind this whole question of validity. > > > > As to how this could proceed, CRISP-DM and KADS may provide some > > pointers, > > albeit they are both European programs. Both started from informal > > working > > groups involving (on the push end) software vendors, consultants or > > academics and (on the pull end) enlightened clients. Ultimately, both > > attracted European Community funding (through the ESPRIT program). In > > fact > > KADS attracted a huge pile of funds - the program has been running > > since > > 1990 and is still underway. > > > > Robert > > > > P.S. Just to give you a flavour of the groups' scope, here are their > > members: > > > > KADS-II (ESPRIT Project P5248) > > Cap Gemini Innovation > > Swedish Institute for Computer Science > > Cap Programator > > Siemens AG > > Netherlands Energy Research Foundation ECN > > Touche Ross Management Consultants > > Entel SA > > University of Amsterdam > > IBM France > > Free University Brussels > > Lloyd's Register > > > > CRISP_DM (ESPRIT Project 24.959) > > Teradata (Denmark) > > SPSS (was ISL) - vendors of 'Clementine' data-mining software > > DaimlerChrysler > > OHRA (Netherlands) > > > > > > > > > > > > > > ========================================================= > > FRIAM Complexity Coffee listserv > > Meets Fridays 9AM @ Museum Hill Cafe > > Archives, unsubscribe, etc.: > > http://www.redfish.com/mailman/listinfo/friam_redfish.com > > > Owen Densmore 451 Camino Don Miguel Santa Fe, NM 87505 > Work: 505-983-6305 Cell: 505-570-0168 Home: 505-988-3787 > [hidden email] http://backspaces.net > http://complexityworkshop.com > > > ========================================================= > FRIAM Complexity Coffee listserv > Meets Fridays 9AM @ Museum Hill Cafe > Archives, unsubscribe, etc.: > http://www.redfish.com/mailman/listinfo/friam_redfish.com > > |
«
Return to Friam
|
1 view|%1 views
| Free forum by Nabble | Edit this page |