tag:blogger.com,1999:blog-2760094876516188582024-02-08T11:05:09.129-08:00CSE574 Planning & LearningClass blog for CSE 574 Planning & Learning (ASU)Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.comBlogger85125tag:blogger.com,1999:blog-276009487651618858.post-11195662602306661272008-09-30T15:32:00.001-07:002008-09-30T15:32:29.328-07:00Readings for link-analysis lectures...<div dir="ltr">We are going back to search-engine technology for next class and will discuss authorities/hubs and page-rank appraoches for page importance. <br><br>The required reading is the relevant sections from<br><br> <a href="http://rakaposhi.eas.asu.edu/cse494/notes/ch2.pdf">http://rakaposhi.eas.asu.edu/cse494/notes/ch2.pdf</a><br><br>you are also welcome to read the optional readings or Kleinberg's JACM paper and Page/Brin/Motwani's unplublished pagerank paper.<br> <br><br>rao<br><br></div> Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-43771950890808193092008-05-29T23:47:00.001-07:002008-05-29T23:47:54.661-07:00Class evaluations for your (temporary) edification..Folks<br><br> It is my custom to share the college teaching evaluations with the students.<br><br>You can look at a copy of the evaluation for 574<br><br><a href="http://rakaposhi.eas.asu.edu/tmp/574-s08.htm">http://rakaposhi.eas.asu.edu/tmp/574-s08.htm</a><br> <br>(The link will be available for about a week). <br><br>Thanks to the 8 of you who took time to do the evaluation (the other four would hopefully be riddled with relentless guilt ;-)<br><br>I appreciate all the comments. <br> <br>regards<br>Rao<br><br> Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-85777103422519203652008-05-13T19:32:00.001-07:002008-05-13T19:32:58.974-07:00The Elusive cse574 Cumulative Grade BookFolks<br><br> Here are the "grades" for the first project, the two homeworks, the final and the all important paper.<br><br> Note that for the paper, I gave the grades in four objectives:<br><br> -->"Significance" -- Is the problem being attempted significant? <br><br> --> "effort" -- How much work did the student put in to the project <br><br> --> "Presentation" -- How readable/compelling is the paper?<br><br> --> "Finish" -- Is the paper self-contained or half-finished?<br><br> The final and the domain project were graded by me. The homework 1 was graded by Menkes and homework 2 was graded by J. Benton.<br><br> The papers were all read by me as well as Will Cushing and we exchanged views on the significance etc. (This of course took the most time).<br><br> Although there is no single number for semester cumulative, I feel that the student "5297 172" certainly deserves an A+. <br> In keeping with my tradition, that student, if he/she so chooses, suggest me what should be the grades for others.<br><br> The grades should be on the system anytime now.. ;-)<br><br> thanks for your work. I will get in touch with some of you with more elaborate paper comments. All the graded material will be available from<br> my office on Friday (I am away at D.C. the next two days)<br><br> Rao<br> -----<br><br> <img src="cid:.0" width=1188 height=291 alt="Emacs!">Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-44819945004884851722008-05-07T10:14:00.000-07:002008-05-07T10:15:03.942-07:00Here is the link for all the semester project reports...Folks<br><br> The following link takes you to a message on the mail archive which has all the project reports sent to me<br><br><a href="http://rakaposhi.eas.asu.edu/s08-cse574-mailarchive/msg00073.html">http://rakaposhi.eas.asu.edu/s08-cse574-mailarchive/msg00073.html</a><br> <br><br>Rao<br><br> Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-4088292023693334602008-05-02T19:03:00.001-07:002008-05-02T19:03:54.711-07:00an old midterm and some homework solutions..<div>Some of you wanted an example exam. Here is one from last offering (this is a midterm though)</div> <div> </div> <div><a href="http://rakaposhi.eas.asu.edu/cse574/1-574-midterm-soln.doc">http://rakaposhi.eas.asu.edu/cse574/1-574-midterm-soln.doc</a></div> <div> </div> <div> </div> <div>Also, you can get solutions for the first homework here</div> <div> </div> <div><a href="http://rakaposhi.eas.asu.edu/cse574/hw1-f04-solns.pdf">http://rakaposhi.eas.asu.edu/cse574/hw1-f04-solns.pdf</a></div> <div> </div> <div><a href="http://rakaposhi.eas.asu.edu/cse574/hw2-f04-solns.pdf">http://rakaposhi.eas.asu.edu/cse574/hw2-f04-solns.pdf</a></div> <div> </div> <div>regards</div> <div>rao</div> <div> </div> <div> </div> Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-17082592141191249632008-05-01T15:48:00.001-07:002008-05-01T15:48:58.072-07:00your peer reviews.. all tf and no idf...I looked at your peer reviews. All I can say is that you should definitely take the classes from each other-- the lowest grade you will give seems to be 8-10 range :-) <br><br>I will have to re-scale and re-center the numbers to mine patterns from it.. There go my hopes of using just your numbers (although they might make good data for collaborative filtering question in my next fall's class).<br> <br>I am slowly reading through the reports. The good news is that as of now, I like at least two of them.. <br><br>rao<br><br> Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-85661555030681710482008-04-29T19:52:00.001-07:002008-04-29T19:52:33.085-07:00Please reply to this mail with your paper and (optionally) presentation as attachmentsthanks<br>RaoSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-85240760528867369892008-04-29T10:46:00.001-07:002008-04-29T10:46:43.558-07:00bring your presentations on a memory stick..Folks<p> Bring your slides for today on a memory stick. We will load them on<br>the class computer in the beginning so the presentations can be<br>streamlined.<p>thanks<br>raoSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-38446598417093803622008-04-29T08:43:00.000-07:002008-04-29T08:44:01.091-07:00Fwd: Real world blocksworldHere is some information about real world block stacking application ;1/2)<p>rao<p><p>---------- Forwarded message ----------<br>From: William Cushing <william.cushing@gmail.com><br>Date: Tue, Apr 29, 2008 at 8:34 AM<br>Subject: Fwd: Real world blocksworld<br>To: "Plan-Yochan@Parichaalak. Eas. Asu. Edu"<br><plan-yochan@parichaalak.eas.asu.edu><p><br>Indeed, Hector took the picture. "Optimizing the Steel Plate Storage<br>Yard Crane Selection Problem", ICAPS 2007 (DC Poster). I believe a<br>short paper version exists in the proceedings somewhere...<p>picture: <a href="http://picasaweb.google.es/hectorpal/Icaps2007/photo#5116899415236524914">http://picasaweb.google.es/hectorpal/Icaps2007/photo#5116899415236524914</a><p>that group's publication list:<br><a href="http://www2.imm.dtu.dk/pubdb/public/publications.php?cmd=full_view&pubtype=&section=7">http://www2.imm.dtu.dk/pubdb/public/publications.php?cmd=full_view&pubtype=&section=7</a><p><br>A related journal article on blockstacking (no joke, that's one of the<br>keywords):<p><a href="http://www.springerlink.com/content/f60826852715v545/">http://www.springerlink.com/content/f60826852715v545/</a><p>(I attached the pdf if you don't have access to springer)<p>-Will<p><br>---------- Forwarded message ----------<br>From: William Cushing <william.cushing@gmail.com><br> Date: Tue, Oct 2, 2007 at 7:07 PM<br>Subject: Real world blocksworld<br>To: plan-yochan <plan-yochan@parichaalak.eas.asu.edu><p><p><a href="http://picasaweb.google.es/hectorpal/Icaps2007/photo#5116899415236524914">http://picasaweb.google.es/hectorpal/Icaps2007/photo#5116899415236524914</a><p>This poster is on a domain which is very blocksworld-ish. Also has<br>towers-of-hanois-ish in that there is no table of infinite capacity;<br>only stacks. However, there appear to be no stacking restrictions<br>(objects don't have different weights or sizes that require sorted<br>stacks).<p>The steel plates *do* have individual names which distinguish them, a<br>highly important technical point (in the comparison with blocksworld).<p>The goal state is not a stacking configuration; the goal is to<br>minimize movement over time (as I understand it), each day one is to<br>deliver certain plates to a conveyer at the far right of the<br>coordinate system. One has a nominal schedule when plates will<br>arrive and when they will be requested, allowing one to subgoal on<br>intermediate stacking configurations (or rather a sequence of stacking<br>configurations) which minimizes movements while achieving those<br>schedules. (and then there is some online planning for each day as<br>requests and supplies deviate from nominal).<p>So many interesting components that bar applying FF directly to the<br>problem...but....a big core part of the technical problem *is*<br>blocksworld. On ~1000 "blocks", in case anyone likes to think of<br>planning + scaling.<p>-WillSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-19593512508272252642008-04-28T11:32:00.001-07:002008-04-28T11:32:33.563-07:00Correction for a homework 2 problem..In Question 1, where I say "Y can start at least 1 unit or at most 2<br>units after X" , please replace the "or" with an "and" (the "or"<br>constraint becomes a non-constraint since you will basically have two<br>constraints [-inf 2] [1 +inf] which is equivalent to [-inf +inf]<p>thanks<br>raoSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-68888524819697563912008-04-26T09:44:00.000-07:002008-04-26T09:48:54.250-07:00Assistance or distraction?The agent is quite intelligent and his behavior is close to the optimal, he is doing well. The assistant learns the agent’s goal and behavior, and starts to act. The agent notices that the world became friendlier, but different. It is time for active exploration (which is expensive) by the agent. The exploration changes goal distribution completely; the assistant distorts the world by trying to achieve wrong goals. After awhile the assistant stops to act, the agent finishes exploration (suspecting he had a glitch); and they are back to the old world.<br />It seems that assistance works, if an assistant much smarter than an agent; and an agent accepting the world as it appears, without trying to learn it.Oleg Bakunhttp://www.blogger.com/profile/04714255510115352657noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-40610589315447562092008-04-23T17:01:00.000-07:002008-04-24T11:35:32.244-07:00*Important*: Format of the class of 4/29--you will do short presentations of your project findingsFolks:<p> For 4/29 (i.e., next Tuesday's class), I will take about 10 min to<br>do a "here is what we did" spiel. The rest of the class will be spent<br>with<br>each of you making a 5-min presentation about your project findings.<br>This will also serve as a quick summary of your report (that you will<br>also<br>bring to the class that day). Ideally, you make this presentation as<br>an advertisement to draw people to read your report (you will also<br>give me electronic--<br>pdf and ppt versions) of the report and talk.<p>I think there is a reasonable amount of diversity in the projects that<br>it is worth everyone hearing summaries of what others did).<p>thanks<br>Rao<br>-------------------<br>Subbarao Kambhampati<br><a href="http://rakaposhi.eas.asu.edu/rao.html">http://rakaposhi.eas.asu.edu/rao.html</a>Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-74617951444810029262008-04-23T14:27:00.001-07:002008-04-23T14:27:52.759-07:00(correct URL) Reading for tomorrow (as well as slides and audi...Sorry,<br> the URL for Minh Do's paper is<br><p><a href="http://www2.parc.com/isl/members/minhdo/publications/2008/do.pdf">http://www2.parc.com/isl/members/minhdo/publications/2008/do.pdf</a><p>rao<p><p>---------- Forwarded message ----------<br>From: Subbarao Kambhampati <subbarao2z2@gmail.com><br>Date: Wed, Apr 23, 2008 at 2:26 PM<br>Subject: [CSE574 Planning & Learning] Reading for tomorrow (as well as<br>slides and audi...<br>To: subbarao2z2@gmail.com<p><br>Folks<p> For Minh Do's guest lecture tomorrow, you might want to read the<br>following paper<p><a href="http://www.aaai.org/Conferences/AAAI/2008/aaai08nectar.php">http://www.aaai.org/Conferences/AAAI/2008/aaai08nectar.php</a><p>(it is<br>only 5 pages!)<p>(Additional related papers are on his homepage at<br><a href="http://www2.parc.com/isl/members/minhdo/">http://www2.parc.com/isl/members/minhdo/</a> )<p><br>----------<p>Also, for those of you who missed Biplav Srivastava's talk on web<br>service composition today, his slides<br>and audio are up on the class notes page <a href="http://rakaposhi.eas.asu.edu/cse574">http://rakaposhi.eas.asu.edu/cse574</a><p>see you tomorrow.<p>Rao<p>--<br> Posted By Subbarao Kambhampati to CSE574 Planning & Learning at<br>4/23/2008 02:26:00 PMSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-41463357837823359032008-04-23T14:26:00.001-07:002008-04-23T14:26:10.055-07:00Reading for tomorrow (as well as slides and audio from today's talk)Folks<p> For Minh Do's guest lecture tomorrow, you might want to read the<br>following paper<p><a href="http://www.aaai.org/Conferences/AAAI/2008/aaai08nectar.php">http://www.aaai.org/Conferences/AAAI/2008/aaai08nectar.php</a><p>(it is<br>only 5 pages!)<p>(Additional related papers are on his homepage at<br><a href="http://www2.parc.com/isl/members/minhdo/">http://www2.parc.com/isl/members/minhdo/</a> )<p><br>----------<p>Also, for those of you who missed Biplav Srivastava's talk on web<br>service composition today, his slides<br>and audio are up on the class notes page <a href="http://rakaposhi.eas.asu.edu/cse574">http://rakaposhi.eas.asu.edu/cse574</a><p>see you tomorrow.<p>RaoSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-29437645521462104232008-04-23T10:14:00.001-07:002008-04-23T10:14:37.781-07:00Fwd: teaching evaluationsYou are so encouraged! As usual, in addition to the numbers, it will<br>be helpful to have<br>written comments on what worked and what didn't.<p>thanks<br>rao<p><p>---------- Forwarded message ----------<br>From: James Collofello <JAMES.COLLOFELLO@asu.edu><br>Date: Wed, Apr 23, 2008 at 9:59 AM<br>Subject: teaching evaluations<br>To: "DL.WG.CEAS.Faculty" <DL.WG.CEAS.Faculty@mainex1.asu.edu><br>Cc: Ann Zell <ann.zell@asu.edu><p><p><p><br>Colleagues,<p><p>The Spring 2008 teaching evaluations became available to students on<br>Mon 4/21, around 10:30 a.m. and will close at Wed 4/30 (reading day)<br>at 12:00 midnight. Students will be able to access the evaluation<br>tool at: <a href="https://fultonapps.asu.edu/eval">https://fultonapps.asu.edu/eval</a><p>Please encourage your students to complete the evaluations, especially<br>the students who enjoy your class, or they will face several nagging<br>email requests.<p><p>jim<p><p>James S. Collofello<p>Associate Dean for Academic and Student Affairs<p>Ira A. Fulton School of EngineeringSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-55267433004477159052008-04-23T09:22:00.001-07:002008-04-23T09:22:58.778-07:00Refreshed the plan recognition slides from the first plan recognition lecture..Folks<p> FYI, I revised and expanded the plan recognition slides from the<br>first lecture based on the actual discussion in the class.<p>raoSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-61858559417876511122008-04-23T06:59:00.001-07:002008-04-23T06:59:24.099-07:00Reminder: Talk Today 11AM on "The Myth and Reality of Web Services Composition" by Biplav Srivastava of IBM research 4/23 11AM BY 510Title:<br> The Myth and Reality of Web Services Composition<p> Speaker:<br> Dr. Biplav Srivastava<br> IBM Research<p> Time: 11AM--12AM; Wednesday April 23rd; BY 510<p> Abstract:<p> Divide-and-conquer or working with complex systems from their<br> basic building blocks is one of the basic tenets of modern engineering.<br> While its applicability to Information Technology has always been<br> felt – example Object Oriented Methodology, its success has been<br> limited. There is a resurgence of interest in componentization of<br> IT systems and services through focus on Service Oriented Architecture,<br> and Web Services as its most popular form. Consequently, Web Services<br> has received wide attention in both academia and IT industry over<br> the past 5-7 years. The attractiveness of this technology lies in<br> the fact that the specifications of the building<br> blocks (i.e., services) are openly available in a registry and<br> so are the building blocks themselves. So, the promise is that a user<br> can build (or modify) an application by composing (or re-composing)<br> components whose specification it discovers from the registry<br> and whose capabilities it can access whenever needed. Depending on<br> what is defined as a service, web services composition can enable<br> many IT issues -- Mashups, Asset Reuse, Business-to-IT alignment,<br> Business-to-Business and Enterprise Application Integration, ...<p> In this talk, we will look at where the hardness of automatically<br> composing web services comes from in practice and how traditional<br> Computer Science techniques (notably planning) have fared. While<br> the original myth was that composition would be hard, in reality,<br> most composition scenarios did not demand scalability of the<br> top-of-the-line planning algorithms. However, what has turned<br> out to be harder than composition is how to set up the composition<br> problem as a traditional Computer Science (notably planning) problem.<br> Two trends are emerging to address this: the composition problem is<br> often cast as a plan reuse and modification problem in the context of<br> richer domain models (e.g., Industry Business Processes), and new<br> composition/ planning paradigms like model-lite planning which are<br> resilient to impoverished domain models.<p> -----------<p> About Biplav:<p> Dr. Biplav Srivastava is a Research Staff Member at IBM Research since<br> February 2001. Though based at IBM's India Research Laboratory, Biplav<br> is on assignment to IBM's T.J.Watson Research Center in Hawthorne, NY, USA.<br> Biplav's research interests are in planning, scheduling, policies,<br> learning and information management, and their practical applications<br> in services -- infrastructure and software (web services), semantic web,<br> autonomic computing and societal domains. Prior to IBM Research, he<br> was Core Technology Architect at an erstwhile Silicon Valley<br> start-up, Bodha, eventually acquired by SAP (2000-2001; process integration),<br> Staff Software Engineer at VLSI/ Philips Semiconductors (1996-2000;<br> electornic design automation) and Assistant System Analyst at TCS,<br> India (1993-1994).Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-85884847221166258922008-04-22T17:57:00.001-07:002008-04-22T17:57:17.718-07:00Aiming to minimize the expected fallout from the unexpected hanging... ;-)I will accept one page (printed) on your review/insight on the<br>decision theoretic assistance paper<br>at the *beginning* of thursday's class. Credit for saying things that<br>I didn't particularly bring out.<p><br>cheers<br>RaoSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-28655269066068095942008-04-22T06:18:00.001-07:002008-04-22T06:18:34.485-07:00Talk on "The Myth and Reality of Web Services Composition" by Biplav Srivastava of IBM research 4/23 11AM BY 510[CSE574 Folks:<p> This is the first of the two application talks planned for this week.<br>The second one will be in class on Thursday. Please make every effort<br>to attend. -Rao]<p><p><br> Title:<br> The Myth and Reality of Web Services Composition<p> Speaker:<br> Dr. Biplav Srivastava<br> IBM Research<p> Time: 11AM--12AM; Wednesday April 23rd; BY 510<p> Abstract:<p> Divide-and-conquer or working with complex systems from their<br> basic building blocks is one of the basic tenets of modern engineering.<br> While its applicability to Information Technology has always been<br> felt – example Object Oriented Methodology, its success has been<br> limited. There is a resurgence of interest in componentization of<br> IT systems and services through focus on Service Oriented Architecture,<br> and Web Services as its most popular form. Consequently, Web Services<br> has received wide attention in both academia and IT industry over<br> the past 5-7 years. The attractiveness of this technology lies in<br> the fact that the specifications of the building<br> blocks (i.e., services) are openly available in a registry and<br> so are the building blocks themselves. So, the promise is that a user<br> can build (or modify) an application by composing (or re-composing)<br> components whose specification it discovers from the registry<br> and whose capabilities it can access whenever needed. Depending on<br> what is defined as a service, web services composition can enable<br> many IT issues -- Mashups, Asset Reuse, Business-to-IT alignment,<br> Business-to-Business and Enterprise Application Integration, ...<p> In this talk, we will look at where the hardness of automatically<br> composing web services comes from in practice and how traditional<br> Computer Science techniques (notably planning) have fared. While<br> the original myth was that composition would be hard, in reality,<br> most composition scenarios did not demand scalability of the<br> top-of-the-line planning algorithms. However, what has turned<br> out to be harder than composition is how to set up the composition<br> problem as a traditional Computer Science (notably planning) problem.<br> Two trends are emerging to address this: the composition problem is<br> often cast as a plan reuse and modification problem in the context of<br> richer domain models (e.g., Industry Business Processes), and new<br> composition/ planning paradigms like model-lite planning which are<br> resilient to impoverished domain models.<p> -----------<p> About Biplav:<p> Dr. Biplav Srivastava is a Research Staff Member at IBM Research since<br> February 2001. Though based at IBM's India Research Laboratory, Biplav<br> is on assignment to IBM's T.J.Watson Research Center in Hawthorne, NY, USA.<br> Biplav's research interests are in planning, scheduling, policies,<br> learning and information management, and their practical applications<br> in services -- infrastructure and software (web services), semantic web,<br> autonomic computing and societal domains. Prior to IBM Research, he<br> was Core Technology Architect at an erstwhile Silicon Valley<br> start-up, Bodha, eventually acquired by SAP (2000-2001; process integration),<br> Staff Software Engineer at VLSI/ Philips Semiconductors (1996-2000;<br> electornic design automation) and Assistant System Analyst at TCS,<br> India (1993-1994).Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-84269357279927517942008-04-22T06:16:00.000-07:002008-04-22T06:17:15.138-07:00talk on Online Continual Planning to Control Modular Production Printer (Minh Do; PARC Labs) 4/24 3:15pm BY 190Title: Online Continual Planning to control Modular Production Printer<p>Speaker: Minh B. Do, (Xerox) PARC Labs, Palo Alto<p>Date/Time: Thursday 4/.24 3:15pm BY 190<p>Abstract: This talk summarizes the recent work at the Embedded<br>Reasoning System at PARC on applying automated planning techniques to<br>the control of modular production printing equipments. These<br>reconfigurable printers radically change the traditional design by<br>using simpler, interchangeable, but smarter components. Like many<br>other real-world applications, such as mobile robotics, this complex<br>domain requires real-time autonomous decision-making and robust<br>continual operation. To our knowledge, this work represents the first<br>successful industrial application of embedded domain-independent<br>temporal planning. Main challenges of applying automated planning<br>technology in this domain include compositional modeling, on-line<br>planning and exception handling, real-time planner control, and the<br>interaction with low-level controller. At the heart of our system is<br>an on-line algorithm that combines techniques from state-space<br>planning and partial-order scheduling. For example, our<br>planning-graph-based planning heuristic takes resource contention into<br>account when estimating makespan remaining. We suggest that this<br>general architecture may prove useful as more intelligent systems<br>operate in continual, online settings. Our system has been used to<br>drive several commercial prototypes and numerous hypothetical (but<br>realistic) printer configurations. When compared with<br>competition-winning state-of-the-art off-line planners in this domain,<br>our system is hundreds of times faster and often finds much better<br>quality plans. At the end of the talk, I will also discuss current<br>extensions of our current planning framework to other online planning<br>domains that share similar characteristics and also to objective<br>functions beyond the default maximization for machine productivity.<p><p>Bio: Minh Do is a Research Staff in the Embedded Reasoning Area at the<br>Palo Alto Research Center (formerly Xerox PARC). He graduated from the<br>Yochan planning group at Arizona State University in 2004 and has been<br>working on transferring his knowledge in offline domain-independent<br>metric temporal planning into fast online continual planning<br>applications. Besides temporal and online planning, Minh Do has<br>worked on other planning topics such as over-subscription planning,<br>planning as CSP/ILP/SAT, and integrating planning and diagnosis. He<br>has published a few dozens papers, filed several patents on automated<br>planning and co-authored the ICAPS best application paper award on<br>planning for high-speed modular printer control. This year, he is<br>co-chairing the deterministic track of the 6th International Planning<br>Competition.Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-19231100178084823302008-04-21T14:01:00.001-07:002008-04-21T14:01:15.083-07:00Re: [CSE574 Planning & Learning] readings for next classshould work now<p>rao<p><br>On Mon, Apr 21, 2008 at 1:50 PM, Aishwarya Sivaraman <asivaram@asu.edu> wrote:<br>> Dr. Rao,<br>><br>> I am unable to open the longer versions' link. I get a 404 error.<br>><br>> <a href="http://rakaposhi.eas.asu.edu/cse574/notes/asst.pdf">http://rakaposhi.eas.asu.edu/cse574/notes/asst.pdf</a><br>><br>><br>> Regards,<br>> Aishwarya<br>><br>><br>> On Mon, Apr 21, 2008 at 1:23 PM, Subbarao Kambhampati<br>> <SUBBARAO.KAMBHAMPATI@asu.edu> wrote:<br>><br>> > For the decision theoretic assistant paper, if you find the short one<br>> > too succinct, you might<br>> > consider the longer version here<br>> ><br>> > <a href="http://rakaposhi.eas.asu.edu/cse574/notes/asst.pdf">http://rakaposhi.eas.asu.edu/cse574/notes/asst.pdf</a><br>> ><br>> > rao<br>> ><br>> ><br>> > On Sun, Apr 20, 2008 at 9:32 PM, Subbarao Kambhampati<br>> > <subbarao2z2@gmail.com> wrote:<br>> > > Folks<br>> > > Here are two readings for next class:<br>> > ><br>> > > Primary: <a href="http://www.cs.orst.edu/%7Eafern/papers/ijcai07-assistant.pdf">http://www.cs.orst.edu/%7Eafern/papers/ijcai07-assistant.pdf</a><br>> > ><br>> > ><br>> > > Secondary:<br>> > ><br>> > > <a href="http://www.cs.rochester.edu/~kautz/papers/gps-tracking.pdf">http://www.cs.rochester.edu/~kautz/papers/gps-tracking.pdf</a><br>> > ><br>> > ><br>> > > Rao<br>> > ><br>> > > --<br>> > > Posted By Subbarao Kambhampati to CSE574 Planning & Learning at<br>> 4/20/2008<br>> > > 09:32:00 PM<br>> > ><br>> > ><br>> > ><br>> > ><br>> > ><br>> > ><br>> ><br>><br>>Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-10345249966140519092008-04-21T13:23:00.001-07:002008-04-21T13:23:36.196-07:00Re: [CSE574 Planning & Learning] readings for next classFor the decision theoretic assistant paper, if you find the short one<br>too succinct, you might<br>consider the longer version here<p><a href="http://rakaposhi.eas.asu.edu/cse574/notes/asst.pdf">http://rakaposhi.eas.asu.edu/cse574/notes/asst.pdf</a><p>rao<p><br>On Sun, Apr 20, 2008 at 9:32 PM, Subbarao Kambhampati<br><subbarao2z2@gmail.com> wrote:<br>> Folks<br>> Here are two readings for next class:<br>><br>> Primary: <a href="http://www.cs.orst.edu/%7Eafern/papers/ijcai07-assistant.pdf">http://www.cs.orst.edu/%7Eafern/papers/ijcai07-assistant.pdf</a><br>><br>><br>> Secondary:<br>><br>> <a href="http://www.cs.rochester.edu/~kautz/papers/gps-tracking.pdf">http://www.cs.rochester.edu/~kautz/papers/gps-tracking.pdf</a><br>><br>><br>> Rao<br>><br>> --<br>> Posted By Subbarao Kambhampati to CSE574 Planning & Learning at 4/20/2008<br>> 09:32:00 PM<br>><br>><br>><br>><br>><br>>Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-87808074759478216722008-04-20T21:32:00.001-07:002008-04-20T21:32:38.718-07:00readings for next classFolks<br> Here are two readings for next class:<p>Primary: <a href="http://www.cs.orst.edu/%7Eafern/papers/ijcai07-assistant.pdf">http://www.cs.orst.edu/%7Eafern/papers/ijcai07-assistant.pdf</a><p><br>Secondary:<p><a href="http://www.cs.rochester.edu/~kautz/papers/gps-tracking.pdf">http://www.cs.rochester.edu/~kautz/papers/gps-tracking.pdf</a><p><br>RaoSubbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-18737703327597760562008-04-17T20:28:00.001-07:002008-04-17T20:28:27.645-07:00Link to the goal graph based plan recognition discussed in today's class..is here<br><a href="http://www.cs.cmu.edu/afs/cs/project/jair/pub/volume15/hong01a.pdf">http://www.cs.cmu.edu/afs/cs/project/jair/pub/volume15/hong01a.pdf</a><br><br>Rao<br><br>ps: Here is a paper talking about minimizing a possibly non-minimal plan. It shows the NP-completeness of <br> minimization and talks about polynomial sound-but-incomplete minimization algorithms<br><br><a href="http://citeseer.ist.psu.edu/fink92spectrum.html">http://citeseer.ist.psu.edu/fink92spectrum.html</a> <br> Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0tag:blogger.com,1999:blog-276009487651618858.post-53515125404003858952008-04-16T09:15:00.001-07:002008-04-16T09:15:40.572-07:00an efficient temporal planner for handling required concurrency..Here is a planner to appear in AAAI that realizes Tempo (described in the temporal planning paper you read) and<br>shows that it is efficient (recall that VHPOP like planners could already solve RC problems--while<br>DEP planners couldn't. Tempo was supposed to be a way of combining reachability heuristics with non-DEP<br> search space. this planner--Crikey3--realizes Tempo by making several interesting improvements to <br>SAPA's relaxed plan graph heuristics). <br><br>FYI<br>rao<br><br><br><a href="http://www.cis.strath.ac.uk/cis/research/publications/papers/strath_cis_publication_2248.pdf">http://www.cis.strath.ac.uk/cis/research/publications/papers/strath_cis_publication_2248.pdf</a><br> Subbarao Kambhampatihttp://www.blogger.com/profile/08449853328445416609noreply@blogger.com0