Machine Learning List: Vol. 13, No. 1 Monday, Feb 26, 2001 Contents Calls for Papers and Other Meeting Announcements CFP: Special Session on "Learning and Adapting in AI Planning" CfP: Machine Learning for User Modeling WS at the UM-2001 UM01 WORKSHOP on PERSONALIZED TELEVISION SERVICES - First CFP JMLR: Special Issue on Kernel Methods Submission to List CFP: IJCAI-01 Workshop - KDD Enhancement Wrappers KDD-2001 Call for Tutorial Proposals UAI-01 Call for Papers cfp: IJCAI-01 Wrkshp on Knowledge Discovery from Heterogeneous, .... CFP: AAAI 2001 Fall Symposium on Using Uncertainity Within Computation IJCAI-01 workshop: Adaptive Text Extraction & Mining IDAMAP 2001 Call for papers CFP: Workshop on Logic & Learning Intl Conf on Comp Intelligence for Modelling, Control and Automation (CIMCA'2001) Jobs postdoc position: U. of Toronto job opportunities in Machine Learning Machine Learning Positions at the Australian National University Other new book on sequence learning The Machine Learning List is moderated. Contributions should be relevant to the scientific study of machine learning. Please send submissions for distribution to: ml@isle.org. For requests to be added, removed, or to change your email address, send email to: ml-request@isle.org. In general, submissions should be no more than a few full-screens of text. For meeting announcements, highlight the conference or workshop web page and give a summary description of the goals of the event. Information such as the list of program committee members, talk schedules, and registration forms are unnecessary and should not be included. Job adds are usually no more than a few full-screens so they should fit naturally. ---------------------------------------------------------------------- Calls for Papers and other Meeting Announcements ------------------------------ From: M.Garagnani@open.ac.uk Subject: CFP: Special Session on "Learning and Adapting in AI Planning" Date: Thu, 25 Jan 2001 15:36:08 +0100 (MET) Preliminary Call for Papers Special Session on "Learning and Adapting in AI Planning" at IC-AI 2001 Las Vegas, Nevada, USA, June 25-28, 2001 http://mcs.open.ac.uk/mg343/AI-session.htm A Special Session on "Learning and Adapting in AI Planning" will take place at the Monte Carlo Resort in Las Vegas, Nevada (USA) during June 25-28, 2001 as part of the 2001 International Conference on Artificial Intelligence (IC-AI 2001, held in conjunction with the International Multiconference: http://www.ashland.edu/~iajwa/conferences/). Scope and Overview: The past few years have seen dramatic advances in planning algorithms and paradigms. Recent systems can quickly solve problems that are orders of magnitude harder than those tackled by the best previous planners. However, if planning systems are to find wide application in real-world situations, they need to be able to offer good quality, real-time performances over a large range of problems. The thesis underlying this session is that in order to do so, not only must planners be fast, but also be flexible and able to adapt automatically to different problems and domains. Two current approaches tackling this issue from different perspectives are: 1) the automatic extraction of domain-specific knowledge through domain analysis, and 2) the acquisition of domain-specific or search-control knowledge and procedural abstractions through the employment of learning techniques. In order to discuss the state of the art of these areas of research and stimulate their cross-fertilisation, we invite submissions of papers on one or both of these approaches. More specifically, topics of particular interest concern the development of planning systems or algorithms that can learn from past experience, and/or automatically analyse and adapt to new problem domains, in order to 1) offer good performances over a wide range of different situations or 2) improve efficiency or plan quality on the basis of failures and successes. Topics will include (but not be limited to): - inference of invariants through domain analysis and hypothesis testing; - acquisition of domain knowledge in incomplete or inaccurate domains; - acquisition of strategies for action selection; - acquisition of heuristics for state-space or plan-space search; - learning macro-operators; - learning plan-rewriting rules; - reinforcement learning for conformant and stochastic planning; - integration of planning, learning and execution; and, in general, the use of any learning technique (such as explanation-based, inductive, supervised, analytic or by-analogy learning) for improvement and optimisation of current state-of-the-art planning systems. Submission Guidelines: [see web page] Important Dates: - March 1, 2001 (Thursday): Draft papers (about 4 to 5 pages) due - April 1, 2001 (Sunday): Notification of acceptance - May 1, 2001 (Tuesday): Camera-Ready papers and Pre-registration due - June 25 - 28, 2001: Session & IC-AI Conference ------------------------------ From: Ralph Schaefer Subject: CfP: Machine Learning for User Modeling WS at the UM-2001 Date: Thu, 25 Jan 2001 15:36:10 +0100 (MET) Call for participation Machine Learning for User Modeling UM-2001 Workshop (http://www.dfki.de/um2001/) User model acquisition is a difficult problem. The information available to a user modeling system is usually limited, and it is hard to infer assumptions about the user that are strong enough to justify non-trivial conclusions. Classical acquisition methods like user interviews, application-specific heuristics, and stereotypical inferences often are inflexible and unsatisfying. The goal of the workshop is twofold: On the one hand, it attempts to be a forum for user modeling researchers who want to discuss specific problems of using machine learning for user modeling. Both experts and novices (and all those in between) are invited. On the other hand, the workshop shall function as a SIG meeting, where joint activities of interested attendants can be planned. Hence, there are two groups of questions to be discussed at the workshop: Research issues: * What learning tasks can be identified in user modeling systems? * Are there classes of problems in user modeling that are particularly well or poorly suited to the application of machine learning methods? * Are there machine learning algorithms or classes of algorithms that are particularly appropriate / not appropriate for user modeling systems? * Are there subareas of user modeling or classes of user modeling systems where machine learning can be especially useful? * In what respects does the induction of a user model differ from other induction tasks to which machine learning is typically applied, and what implications does this have for the application of machine learning in user modeling? Participation and Paper Submission: Participants are required to submit a short paper that - describes why they are interested in the application of machine learning techniques to user modeling and the problems and questions they have encountered and/or - makes proposals concerning SIG activities and/or - describe their current work and interests as related to the workshop topic In the first two cases, authors shall provide comments and answers to the questions above as topics of interest, and perhaps raise new relevant questions and issues in about 2 pages. In the third case, the work and interests should be described in no more than 10 pages. Participants will be selected based on their submissions. Submission instructions: Please submit a short paper in PostScript, PDF, or HTML to Ralph.Schaefer@dfki.de. The final version should not exceed 10 pages. Deadlines: March 8 deadline for submissions April 1 notification of authors about acceptance April 27 deadline for revised versions of accepted contributions May 11 accepted contributions and first draft of the workshop program made available to participants; mailing list for participants set up ------------------------------ From: Liliana Ardissono Subject: UM01 WORKSHOP on PERSONALIZED TELEVISION SERVICES - First CFP Date: Wed, 03 Jan 2001 12:47:48 +0100 CALL FOR PAPERS International Conference on User Modeling http://www.dfki.de/um2001/ July 13 to July 17, 2001 AlpenCongressCentrum in Sonthofen, Bavaria, Germany Workshop on Personalization in Future TV July 13-14, 2001 With the advent of digital networks, the world of TV as we know it -- mass-media broadcast -- is undergoing tremendous changes. The increase of the number of available channels, the convergence of TV and internet, and the proliferation of new interactive services will transform the TV box function from a program watching device to a portal towards all kinds of content and services. The next TV era will revolve around sophisticated set-top boxes integrating viewing, listening and recording functionalities, connections to several sources (internet, cable, satellite), games, as well as communications features. In order to cope with the complexity of such an environment and efficiently choose among the huge amount of available alternatives, the users are in need of an advanced user interface to provide them with an intelligent assistance. In particular, personalization is taking an increasingly important role in the design of adaptive user interfaces, which can focus on the services interesting to the user and tailor the suggestion of the available options to her interests. The aim of this workshop is to provide a forum in which researchers from diverse fields such as machine learning, knowledge engineering, psychology, cognitive sciences, adaptive user interfaces, user modeling and business intelligence can examine the personalization aspects of the user interface in future TV. Topics to be addressed include, but are not limited to: - Position papers: new interactive services, social and business implications - Applications: TV portal architectures, personal TV recommender systems - Content recommendation: new architectures for recommender systems, cross-domain recommendation - Adaptive presentation: adaptive hypermedia techniques for EPGs, EPGs for all: adaptivity to users with special needs - Interaction: new browsing paradigms, conversational interfaces - Management of the user model - Context: single user vs. multiple users, context detection and integration (mood, time, location, ...) - Community formation and social influences SUBMISSION INSTRUCTIONS: All papers must include in the first page: the title, author's name(s), affiliation, mailing address, phone number, e-mail, home page URL, and up to five keywords. Full papers must also include an abstract of 200 words maximum. Papers should be prepared as PostScript or PDF files, written in Times, 11pt and printable on A4 paper or US Letter. If possible, the papers should be formatted according to the guidelines used for the papers of the main conference (see http://www.dfki.de/um2001/). Suitable templates (LaTex2e | LaTex | Tex | MS Word (PC) | MS Word (Mac)) can be retrieved from the LNCS Web site at the following URL: http://www.springer.de/comp/lncs/authors.html. The size of original submission is limited to 4 pages for position papers and 8 pages for full papers. Electronic submission of the URL address of the paper is preferred, although e-mail submissions of the Postscript/PDF files will be accepted. Send your submission to Liliana Ardissono: liliana@di.unito.it. Time Schedule Submission of contributions: March 8, 2001 Notification of acceptance: April 1, 2001 Submission of final contributions: April 20, 2001 ------------------------------ From: Nello Cristianini Subject: JMLR: Special Issue on Kernel Methods Date: Sun, 7 Jan 2001 09:15:33 +0000 (GMT) CALL FOR PAPERS Journal of Machine Learning Research Special Issue on "New Perspectives on Kernel Based Learning Methods" http://www.cs.rhbnc.ac.uk/colt/JMLRspecissue.html Guest Editors: Nello Cristianini, John Shawe-Taylor, Bob Williamson Important dates: Submission deadline: March 15th, 2001 Decision : May 15th, 2001 Final Versions : June 15th, 2001 Submission procedure: see webpage: http://www.cs.rhbnc.ac.uk/colt/JMLRspecissue.html Background: Recent theoretical advances and experimental results have drawn considerable attention to the use of kernel functions in learning systems. Support Vector Machines, Gaussian Processes, kernel PCA, kernel Gram-Schmidt, Bayes Point Machines, Relevance and Leverage Vector Machines, are just some of the algorithms that make crucial use of kernels for problems of classification, regression, density estimation, novelty detection and clustering. At the same time as these algorithms have been under development, novel techniques specifically designed for kernel-based systems have resulted in methods for assessing generalisation, implementing model selection, and analysing performance. The choice of model may be simply determined by parameters of the kernel, as for example the width of a Gaussian kernel. More recently, however, methods for designing and combining kernels have created a toolkit of options for choosing a kernel in a particular application. These methods have extended the applicability of the techniques beyond the natural Euclidean spaces to more general discrete structures. The field is witnessing growth on a number of fronts, with the publication of books, editing of special issues, organization of special sessions and web-sites. Moreover, a convergence of ideas and concepts from different disciplines is occurring. This special issue will accept papers in any of the following main research directions: 1) design of novel kernel-based algorithms 2) design of novel types of kernel functions 3) development of new learning theory concepts 4) application of the techniques to new problem areas More information at: http://www.cs.rhbnc.ac.uk/colt/JMLRspecissue.html Or: nello@dcs.rhbnc.ac.uk ------------------------------ From: "Sarabjot S. Anand" Subject: Submission to List Date: Sun, 7 Jan 2001 14:00:11 -0000 Call for Papers IJCAI-01 Workshop on Intelligent Techniques for Web Personalization (ITWP'2001) Seattle, Washington, USA 4-6 August 2001 http://www.mineit.com/lc/ijcai Overview -------- The intense competition among Internet-based businesses to acquire new customers and retain the existing ones has made Web personalization an indispensable part of e-commerce. Web personalization can be defined as any action that tailors the Web experience to a particular user, or set of users. To achieve effective personalization, organizations must rely on all available data, including the usage and click-stream data, the site content, the site structure, as well as user demographics and profiles. In addition, efficient and intelligent techniques are needed to mine this data for actionable knowledge, and to effectively use the discovered knowledge to enhance the users' Web experience. These techniques must address important challenges emanating from the size and the heterogeneous nature of the data itself, as well as the dynamic nature of user interactions with Web sites, especially in e-commerce applications. These challenges include the scalability of the personalization solutions, data integration, and successful integration of techniques from machine learning, information retrieval and filtering, databases, data mining, text mining, statistics, and human-computer interaction. Topics ------ The following are some of the issues and topics that will be addressed at the workshop: Data and Knowledge Modelling, Integration and Management o Data models for Web usage, content, and structure data o Integration of content, structure and usage data for personalization o Techniques for improving online data quality o The role of cross channel data in online personalisation o Knowledge representation issues Architectures and Systems o Scalable collaborative filtering techniques o Effective techniques for personalization based on anonymous data o Agents for intelligent browsing and navigation o Automated techniques for generation and updating of user profiles o Cognitive models for Web navigation and e-commerce interactions o Metrics for Personalisation Effectiveness Enabling Technologies for Personalisation o Adaptive hypertext systems o Personalization based on Web usage mining o Text mining techniques for content-based filtering Paper Format, Submission and Publication [see web page] ---------------------------------------- Important Dates --------------- Abstract Submission: 19 February, 2001 Paper Submission: 26 February, 2001 Notification of Acceptance: 30 March, 2001 Camera Ready Paper Due: 25 April, 2001 Worksop: 4 August, 2001 ------------------------------ From: "William H. Hsu" Subject: CFP: IJCAI-01 Workshop - KDD Enhancement Wrappers Date: Mon, 8 Jan 2001 09:30:32 -0600 FIRST CALL FOR PARTICIPATION IJCAI-2001 Workshop on Wrappers for Performance Enhancement in Knowledge Discovery in Databases (KDD) [workshop code ML-5] http://www.kddresearch.org/KDD/Workshops/IJCAI-2001/ Saturday, 4 Aug 2001 Seattle, Washington, USA WORKSHOP DESCRIPTION The rapidly increasing volume of data collected for decision support applications in commercial, industrial, medical, and defense domains has made it a challenge to scale up knowledge discovery in databases (KDD), the machine learning and knowledge acquisition component of these applications. Many techniques currently applied to KDD admit enhancement through the WRAPPER approach, which uses empirical performance of inductive learning algorithms as feedback to optimize parameters of the learning system. Wrappers include algorithms for performance tuning, especially: optimization of learning system parameters (HYPERPARAMETERS) such as learning rates and model priors; control of solution size; and change of problem representation (or inductive bias optimization). Strategies for changing the representation of a machine learning problem include decomposition of learning tasks into more tractable subproblems; feature construction, or synthesis of more salient or useful input variables; and feature subset selection, also known as variable elimination (a form of relevance determination). This workshop will explore current issues concerning wrapper technologies for KDD applications. WORKSHOP AUDIENCE This workshop is intended for researchers in the area of machine learning, including practitioners of knowledge discovery in databases (KDD) and statistical and computational learning theorists. Intelligent systems researchers with an interest in high-performance computation and large-scale, real-world applications of data mining (e.g., inference and decision support) will also find this workshop of interest. CALL FOR PAPERS We encourage submissions containing original theoretical and applied concepts in KDD. Experimental results are also encouraged, especially on fielded applications, even if they are only preliminary. We therefore invite two categories of paper submissions: - research papers - short summaries (including position papers) For the workshop agenda, submission procedure, and up-to-date information on the review committee and invited speakers, please visit the workshop web site: www.kddresearch.org/KDD/Workshops/IJCAI-2001/ IMPORTANT DATES Full Papers due: Friday, 02 Mar 2001 Short Papers due: Friday, 16 Mar 2001 acceptance notification: Friday, 30 Mar 2001 camera-ready copy due: Friday, 13 Mar 2001 workshop Saturday, 04 Aug 2001 ------------------------------ From: Tom Fawcett Subject: KDD-2001 Call for Tutorial Proposals Date: Mon, 8 Jan 2001 15:12:27 -0800 Call for Tutorial Proposals KDD-2001: The Seventh ACM SIGKDD International Conference on Knowledge Discovery and Data Mining August 26-29, 2001, San Francisco, California, USA http://www.acm.org/sigkdd/kdd2001 INTRODUCTION Tutorials have become an integral part of KDD conferences. This is partly because of the interdisciplinary nature of data mining, but also because of the amount and speed of progress in the past decade. Tutorials are an effective way to educate conference attendees in specific topics and emerging sub-fields. Traditionally, KDD conferences have offered high quality tutorials on many aspects of data mining. For KDD-2001, we are seeking proposals for 4 to 8 tutorials, each of 1.5 or 3 hours duration. A tutorial should ideally appeal to more than one sub-community of data mining, i.e., databases, machine learning and statistics. It should both educate conference attendees about a subfield and provide background necessary to understand technical advances. It may discuss novel data mining techniques, successful applications in data mining, and/or theme-oriented comprehensive surveys. PROPOSAL SUBMISSIONS Tutorial proposals should be submitted by March 7, 2001. Proposals should be submitted electronically to the Tutorials Chair (Tom Fawcett ). Proposals may take the form of plain text, Postscript, PDF, Microsoft Word, Microsoft Powerpoint, or some combination of these. Contact the Tutorials Chair if you wish to provide non-electronic supporting materials along with your proposal. PROPOSAL DETAILS Tutorial proposals should address the following issues: (a) Basic information: Title, brief description, names and contact information for each tutor, the length of the proposed tutorial (1.5 or 3 hours). (b) Audience: What is the intended audience for the tutorial, e.g., novice users of statistical techniques, expert researchers in text mining, or database administrators. (c) Interest: Why is this topic important/interesting to the KDD community? Provide some informal evidence that people would attend. Evidence might include related workshops, conference sessions, mailing lists, discussions, papers, symposia, communities, etc. (d) Coverage: How deep/broad is the proposed tutorial? How valuable would the tutorial be with the given scope? (e) Background: What background will be required of the audience? Enough materials should be included in the proposal to provide a sense of the scope and depth of the tutorial. The more details that can be provided, the better; up to and including actual overhead slides. The proposal should also include some biographical information on each tutor (including WWW address, if applicable). This information should describe the qualifications of each presenter with respect to the tutorial's topic. For the proposed subject matter the tutor should have appropriate qualifications. On the other hand, the tutor should NOT focus mainly on his or her research results. A KDD tutorial is not a forum for promoting one's research or product. If, for certain parts of the tutorial, the material comes directly from the tutor's own research or product, please indicate this in the proposal. IMPORTANT DATES March 7, 2001: Tutorial proposals due March 30, 2001: Notification of proposal acceptance August 26, 2001: KDD-2001 Tutorials held For further information, please contact Tom Fawcett (tom_fawcett@hp.com). ------------------------------ From: Daphne Koller Subject: UAI-01 Call for Papers Date: Sun, 14 Jan 2001 12:45:01 -0800 UAI-2001: Call for Papers August 2-5, 2001 University of Washington Seattle, Washington, USA Conference homepage: http://robotics.stanford.edu/~uai01/ Uncertainty management is a key enabling technology for the development of intelligent systems. Since 1985, the Conference on Uncertainty in Artificial Intelligence (UAI) has been the primary international forum for exchanging results on the use of principled uncertain-reasoning methods in intelligent systems. The conference has catalyzed advances in fundamental theory, efficient algorithms, and practical applications. Theory and technology first presented at UAI have been proven by their wide application in the scientific, commercial, and industrial communities. The UAI Proceedings have become a fundamental reference for researchers and practitioners who want to know about both theoretical advances and the latest applied developments in the field. The scope of UAI is wide, covering a broad spectrum of approaches to automated reasoning, learning, decision making and knowledge acquisition under uncertainty. Contributions range from those that that advance theoretical principles to those that provide insights through the empirical study of applications, from quantitative to qualitative approaches, from traditional to non-classical paradigms for uncertain reasoning, and from autonomous systems to those designed to support human decision making. We encourage submissions of papers for UAI-2001 that report on advances in the core areas of representation, inference, learning, decision making, and knowledge acquisition, as well those dealing with on insights derived from the construction and use of applications involving uncertain reasoning. SUBMISSION INFORMATION Deadlines: Abstracts (200 words): Monday, March 12, 2001 (11:59PM PST) Full papers: Tuesday, March 20, 2001 (11:59PM PST) The deadlines will be strictly enforced (the submission server will be closed at midnight). No extensions will be granted under any circumstances. Papers and abstracts should be submitted through http://cmt.research.microsoft.com/UAI2001/ If authors have special circumstances that prevent electronic submission, arrangements can be made directly with the program chairs below. Authors are required to submit papers in the proceedings format. Submitted papers must be no more than eight pages in proceedings format, including figures and bibliography (about 5600 words). Accepted papers will be alloted eight pages in the conference proceedings, with two additional pages available for a fee. Please see http://robotics.stanford.edu/~uai01/FormatInstructions.html for format information and access to style files. Papers submitted for review should represent original, previously unpublished work. Papers should not be under review for presentation in any other conference; however, an extended version of the paper may be under review for publication in a scientific journal. Submitted papers will be carefully evaluated on the basis of originality, significance, technical soundness, and clarity of exposition. Papers may be accepted for presentation in plenary or poster sessions. All accepted papers will be included in the Proceedings of the Seventeenth Conference on Uncertainty in Artificial Intelligence, published by Morgan Kaufmann Publishers. Other important dates: Author Notification of Accepted Papers: April 30, 2001 Camera-ready Copy of Accepted Papers due: June 4, 2001 Workshops and Tutorials: Thursday, August 2, 2001 Technical Program: Friday, August 3-Sunday, August 5, 2001 ------------------------------ From: Vasant Honavar Subject: cfp: IJCAI-01 Wrkshp on Knowledge Discovery from Heterogeneous, .... Date: Mon, 15 Jan 2001 17:40:12 -0600 (CST) Call for Papers IJCAI-01 Workshop on Knowledge Discovery from Distributed, Dynamic, Heterogeneous, Autonomous Data and Knowledge Sources (http://www.cs.iastate.edu/~honavar/ijcai00workshop.html) to be held at the International Joint Conference on Artificial Intelligence (ICAI-2001) August 6, 2001 Seattle, Washington, USA. Background Recent advances in sensor, high throughput data acquisition, and digital information storage technologies, have made it possible to acquire and store large volumes of data in digital form. Advances in computers and communications, the Internet, and mobile computing have made it possible for scientists and decision makers, at least in principle, to access and utilize this data for data-driven knowledge acquisition and decision making in the respective domains. Examples of such domains include bioinformatics, monitoring and control of complex, dynamic, distributed systems (e.g., communication networks, power systems), among others. Despite the diversity of these domains, they share several common characteristics: * Different data sources often provide different types of data (e.g., signals from sensors, relational data, text, images, macromolecular (DNA and protein) sequences, protein structures, simulations). This calls for sophisticated tools for selective and context-sensitive information extraction and information fusion. Such tools have to be able to bridge the gap in structure and semantics of the respective data and knowledge sources (e.g., using domain-specific ontologies). * Data Repositories of interest are physically distributed. Given the large amounts of data that is being gathered and stored at these repositories, and the fact that users are typically interested not in the raw data, but in results of analysis of the data in a given context, it is desirable to process the data in a distributed fashion wherever the data is located and selectively transmit the results of analysis. This calls for efficient and scalable analysis tools (e.g. data mining algorithms and decision making algorithms) with provable performance guarantees in a distributed setting. * Data sources are often autonomous and the nature of access to data that is available is often restricted due to privacy and security considerations. Thus, users have a limited view of the data (e.g., in the form of statistical summaries or results of an agreed-upon set of operations). Thus there is a need for systematic analysis of the information requirements of data analysis or decision making algorithms in such environments. * Data sources are dynamic. Given the large amounts of data that need to be processed, this calls for efficient incremental or cumulative algorithms that can update the results of analysis (e.g., a hypthesis generated by a data mining algorithm). * The goals and consequently information needs of users as well as the data sources can change over time. This calls for development of information extraction and fusion algorithms and data mining algorithms that can dynamically adjust to shifting goals and changing constraints. Topics of Interest We invite full papers, extended abstracts, or position papers on all aspects of knowledge discovery from distributed, dynamic, heterogeneous, autonomous data and knowledge sources. Important Dates and Deadlines: * Deadline for submission of full papers: March 1, 2001. * Deadline for submission of position papers or abstracts: March 15, 2001. * Notification of acceptance: March 30, 2001. * Deadline for receipt of camera-ready papers: April 21, 2001 * Workshop: August 6, 2001. Instructions for Authors [see web page] ------------------------------ From: Ian Miguel Subject: CFP: AAAI 2001 Fall Symposium on Using Uncertainity Within Computation Date: Fri, 19 Jan 2001 08:14:19 +0000 Using Uncertainity Within Computation ===================================== Call for Participation AAAI 2001 Fall Symposium Sea Crest Oceanfront Resort and Conference Center North Falmouth, Cape Cod, MA November 2-4, 2001 http//:www.cs.york.ac.uk/~tw/fall Outline ------- To reason about complex computational systems, researchers are starting to borrow techniques from the field of uncertainty reasoning. In some cases, this is because the algorithms contain stochastic components. For example, Markov decision processes are now being used to model the trajectory of stochastic local search procedures. In other cases, uncertainity is used to help model and cope with the stochastic nature of inputs to (possibly deterministic) algorithms. For example, Monte Carlo sampling is used to deal with uncertainity in game playing programs, whilst probability distributions are used to model variations in runtime performance. Uncertainity and randomness have also been found to be a useful addition to many deterministic algorithms. And a number of areas like planning, constraint satisfaction, and inductive logic programming which have traditionally ignored uncertainity in their computations are waking up to the possibility of incorporating uncertainity into their formalisms. The goal of this workshop is to encourage symbiosis between these different areas. Topics ------ The aim is to bring together researchers from a number different areas of AI including (but not limited to) agents, constraint programming, decision theory, game playing, knowledge representation and reasoning, learning, planning, probabilistic reasoning, qualitative reasoning, reasoning under uncertainty, and search. Possible topics include (but are not limited to): o Incorportating uncertainity into existing frameworks o Modelling uncertainity in computation o Monte Carlo sampling o Probabilistic analysis and evaluation of algorithms o Randomization of algorithms o Stochastic vs. systematic algorithms o Utility and computation Submission Information ---------------------- http://www.cs.york.ac.uk/~tw/fall ------------------------------ From: Nicholas Kushmerick Subject: IJCAI-01 workshop: Adaptive Text Extraction & Mining Date: Fri, 19 Jan 2001 15:32:55 +0000 2nd CALL FOR PAPERS IJCAI-01 Workshop on Adaptive Text Extraction & Mining (ATEM-2001) 5 August 2001 / Seattle, USA in conjunction with the 17th International Joint Conference on Artificial Intelligence http://www.smi.ucd.ie/ATEM2001 papers due 9 March 2001 -- Nicholas Kushmerick \ Fabio Ciravegna | organizing Raymond Mooney | committee Ion Muslea / ------------------------------ From: Riccardo Bellazzi Subject: IDAMAP 2001 Call for papers Date: Mon, 22 Jan 2001 17:05:54 +0100 First Call for Papers INTELLIGENT DATA ANALYSIS IN MEDICINE AND PHARMACOLOGY A Workshop at the Medinfo2001 London, UK, September 4th, 2001 http://magix.fri.uni-lj.si/idamap2001 GENERAL INFORMATION IDAMAP-2001, a one day Medinfo2001 Workshop, will be held in London, UK, on Tuesday, 4th of September, 2001. This is the sixth IDAMAP Workshop: the former ones were held in Budapest in 1996, Nagoya in 1997, Brighton in 1998, Washington DC in 1999, and Berlin in 2000. Gathering in an informal setting, workshop participants will have the opportunity to meet and discuss selected technical topics in an atmosphere which fosters the active exchange of ideas among researchers and practitioners. The workshop is intended to be a genuinely interactive event and not a mini-conference, thus ample time will be allotted for general discussion. TOPIC A large amount of data is currently being collected in bio-medicine for both experimental and observational purposes. This automatic data collection pushes towards the development of methods and tools able to handle and analyze data in a computer-supported fashion, and to support evidence to be exploited in all activities of the biomedical field. In the majority of the application areas, this task cannot be accomplished without using the available knowledge on the domain or on the data analysis process. This need becomes crucial in clinical applications, since medical decision making needs to be supported by arguments based on basic medical and pharmacological knowledge. The topics of the workshop are computational methods for biomedical data analysis that aim to narrow the gap between data gathering and data comprehension and to support the use and exploitation of observational and retrospective data. In terms of methodology, topics include, but are not limited to, - data mining techniques, including machine learning, clustering, neural networks, etc., - other techniques for construction of predictive models, - data visualization, - interpretation of time-ordered data (derivation and revision of temporal trends and other forms of temporal data abstraction), - knowledge management and its integration with intelligent data analysis techniques, - utility of background knowledge in data analysis, - integration of intelligent data analysis techniques within biomedical information systems. DEADLINES May 18, 2001 Papers submitted June 8, 2001 Notification of acceptance June 29, 2001 Final version of accepted papers received SUBMISSION AND PUBLICATION OF PAPERS [see web page] ------------------------------ From: Roni Khardon Subject: CFP: Workshop on Logic & Learning Date: Wed, 24 Jan 2001 18:50:47 -0500 (EST) Call For Papers Workshop on Logic and Learning Affiliated with LICS 2001 June 19-20, 2001, Boston, Massachusetts http://www.eecs.tufts.edu/~roni/LicsWksp/ Logic has been used as the underlying representation language in many areas of AI including machine learning. Learnability of logical expressions has been studied in many paradigms including PAC learning, query based learning, inductive inference, and inductive logic programming. There are theoretical results on learning in propositional logic as well as for logic programs, description logic, and fragments of first-order logic. The techniques applied are probabilistic and combinatorial, recursion theoretic, proof theoretic, and model theoretic. The workshop aims to focus on such logic-based results and techniques for learning, fostering further understanding of the use of logic in learning. The workshop has a two-fold objective: to provide an introduction to the area for those who work in other LICS areas and are interested in applying logic to learning, and to provide a forum for research in the area of logic learning. The workshop will feature invited talks by experts in the field, and contributed talks presenting new research results. Confirmed invited speakers include: Peter Flach (University of Bristol, UK), Lisa Hellerstein (Polytechnic University, USA), Arun Sharma (University of New South Wales, Australia), and Leslie Valiant (Harvard University, USA). We invite submissions of papers presenting new results and/or position papers highlighting particular aspects of logic and learning. No publication of proceedings is planned but accepted papers will be collected in a booklet of workshop notes. A selection of papers from the workshop may be published in a special issue of The Annals of Mathematics and Artificial Intelligence. Participants will be invited to submit full versions for this special issue. Topics of interest include (but are not limited to): inductive logic programming, logical aspects of inductive inference, PAC and related learning models for logic, finite model theory and learning, logical aspects of natural language learning. Submitted papers should be extended abstracts of 2-6 pages in length. Papers should be submitted, preferably in electronic form in Postscript format, to: roni@eecs.tufts.edu. Alternatively, printed copies can be sent to : Roni Khardon EECS Department, Tufts University 161 College Ave., Medford, MA 02155, USA Important dates: submission: 3/15, notification of acceptance: 4/15. Web Info: * Workshop Web Page: http://www.eecs.tufts.edu/~roni/LicsWksp/ * The IEEE Symposium on Logic in Computer Science (LICS) 2001 http://www.cs.bu.edu/faculty/mairson/LICS01/index.html * LICS main web site http://www.math.uic.edu/lics/ ------------------------------ From: CIMCA Subject: Intl Conf on Comp Intelligence for Modelling, Control and Automation (CIMCA'2001) Date: Tue, 30 Jan 2001 15:27:55 +1100 International Conference on Computational Intelligence for Modelling, Control and Automation - CIMCA'2001 9-11 July 2001 Las Vegas, USA http://beth.canberra.edu.au/conferences/CIMCA2001/index.htm In Conjunction with International Conference on Intelligent Agents, Web Technologies and Internet Commerce - iawtic'2001 http://beth.canberra.edu.au/conferences/IAWTIC2001/index.htm CALL FOR PAPERS ======================= Honorary Chairs: Lotfi A. Zadeh, University of California, USA Stephen Grossberg, Boston University, USA The international conference on computational intelligence for modelling, control and automation will be held in Las Vegas, USA on 9-11 July 2001. The conference provides a medium for the exchange of ideas between theoreticians and practitioners to address the important issues in computational intelligence, modelling, control and automation. The conference will consist of both plenary sessions and contributory sessions, focusing on theory, implementation and applications of computational intelligence techniques to modelling, control and automation. For contributory sessions, draft papers (4 pages or more) are being solicited. Several well-known keynote speakers will address the conference. Paper Submission ================ Papers will be selected based on their originality, significance, correctness, and clarity of presentation. Extended abstract (4 pages) should be submitted to the following e-mail or the following address: CIMCA'2001 Secretariat School of Computing University of Canberra Canberra, 2601, ACT, Australia E-mail: cimca@ise.canberra.edu.au E-mail submission is preferred. Extended abstract should present original work, which has not been published or being reviewed for other conferences. Important Dates =============== 16 March 2001 Deadline for submission of draft papers 16 April 2001 Notification of acceptance 16 May 2001 Deadline for camera-ready copies of accepted papers 9-11 July 2001 Conference sessions ------------------------------ Jobs ------------------------------ From: Richard Zemel Subject: postdoc position: U. of Toronto Date: Tue, 9 Jan 2001 15:44:05 -0500 ************************************************************************ POSTDOCTORAL RESEARCH ASSOCIATE DEPARTMENT OF COMPUTER SCIENCE UNIVERSITY OF TORONTO A postdoctoral position is available in the Machine and Neurobiological Learning laboratory in the Department of Computer Science at the University of Toronto. For more details on research in this lab, see http://www.cs.toronto.edu/~zemel/research.html. The University of Toronto is a prominent research institute in all areas of computer science, and a world leader in artificial intelligence. We are seeking someone whose research focus lies primarily in at least one of these areas: machine learning, neural computation, or probabilistic reasoning. A strong background and education in a quantitative discipline, such as physics, mathematics, statistics, engineering, or computer science, is required. Knowledge of neuroscience or psychophysics is a plus. This position is available immediately and for a duration of 1-2 years depending on accomplishment. A competitive salary package will be offered. Please send a CV, a brief statement of research experience and interests, and the names and contact information of two references to: zemel@cs.toronto.edu. Applications can also be mailed (or faxed) to: Dr. Richard Zemel Tel: (416) 978-7497 Departmet of Computer Science Fax: (416) 978-1455 University of Toronto Email: zemel@cs.toronto.edu Toronto, ON M5S 1A4 CANADA URL: www.cs.toronto.edu/~zemel ************************************************************************ ------------------------------ From: "Teller, Astro" Subject: job opportunities in Machine Learning Date: Sun, 14 Jan 2001 12:15:09 -0500 Positions Open on BodyMedia's Advanced Development Team BodyMedia helps people lead a balanced healthy lifestyle through personalized body monitoring. BodyMedia's on-line HealthManager and SenseWear armband monitor allow individuals to track their personal daily health routines, enabling them to play a more proactive role in their wellness management and gain balance in their lives. The Advanced Development Team at BodyMedia consists of machine learning specialists, data miners, computer scientists, signal processing specialists, software engineers and statisticians. This group explores, identifies, and creates new algorithms and products that leverage the rich, continuous, multi-parameter biometric data collected and stored as part of the BodyMedia system. Senior Data Miner: will create algorithms that scale to massive quantities of data. The Senior Data Miner will develop and support data schema, data dictionary, and a data map. This individual will address inter-company inquiries relating to this data, and develop new queries and query procedures regarding the data. Qualifications: Experience in data mining, algorithms, data analysis, data warehousing, Java and C++ Master's degree with 7 to 10 years of related experience, or Ph.D. in Computer Science with 2 to 3 years of related experience. Signal Processing Specialist: will have an understanding of product needs and specifications and how they impact research and development projects. The Signal Processing Specialist will design, implement, and evaluate various innovations related to signal processing. This individual is jointly responsible for the design and implementation of symbolic and numerical algorithms. This individual will also develop real-time signal processing in embedded hardware, offline signal processing, and understanding for pattern recognition, trending, and production purposes. Qualifications: Master's degree with 7 to 10 years of related experience, or Ph.D. in Computer Science or Electrical Engineering with 2 to 3 years of related experience. Senior Machine Learning Specialist: will be familiar with symbolic and numerical learning methods. This individual will also have experience with supervised and unsupervised learning methods, real-time and offline learning methods. The ideal candidate is an even blend of engineer and scientist. Qualifications: Master's degree with 7 to 10 years of related experience, or Ph.D. in Computer Science with 2 to 3 years of related experience. The Director of Informatics and Signal Understanding: will be leading and managing a team of 5 to 6 data miners, machine learning specialists, and computer scientists. This group will focus on advanced development and qualified applicants should have an extensive background in information technology, science methodology, signals, and Artificial Intelligence. Qualifications: Master's degree with 7 to 10 years of related experience, or Ph.D. in Computer Science with 2 to 3 years of experience. Demonstrated excellence in management skills. BodyMedia, Inc. is a Pittsburgh-based, venture-backed company. To contact BodyMedia about these jobs or other open positions at BodyMedia, send email to "jobs@bodymedia.com" To learn more about BodyMedia, visit us on the web at www.bodymedia.com ------------------------------ From: Bob Williamson Subject: Machine Learning Positions at the Australian National University Date: Thu, 1 Feb 2001 13:59:24 +1100 (EST) There are openings for up to 5 academic positions in Computer Science at the Australian National University with a particular interest in candidates with a Machine Learning background. Up to 3 of the positions are tenured. Applications are sought at all levels B-E in the Australian system (=lecturer to Professor in the UK; = Assistant to Full Professor in the US). Details are available at http://www.rsise.anu.edu.au/positions/acad1.html The closing date is 2 March. As well as the contact details provided in web page referred to above, feel free to contact me as well. +------------------------------------------------------------------------+ | Bob Williamson Email: Bob.Williamson@anu.edu.au | | Department of Engineering Phone: +61 2 6125 0079 | | Australian National University Dept.: +61 2 6125 3378 | | Canberra 0200 Fax: +61 2 6125 0506 | | AUSTRALIA http://theorem.anu.edu.au/~williams/home.shtml | +------------------------------------------------------------------------+ ------------------------------ Other ------------------------------ From: Ron Sun Subject: new book on sequence learning Date: Fri, 26 Jan 2001 23:37:31 -0600 SEQUENCE LEARNING: PARADIGMS, ALGORITHMS, AND APPLICATIONS edited by: Ron Sun and C. L. Giles published by Springer-Verlag: LNAI 1828 This book is intended for use by scientists, engineers, and students interested in sequence learning in artificial intelligence, neural networks, and cognitive science. The book will introduce essential algorithms and methods of sequence learning and further develop them in various ways. With the help of these concepts, a variety of applications will be examined. This book will allow the reader to acquire an appreciation of the breadth and variety sequence learning and its potential as an interesting area of research and application. The reader is assumed to have basic knowledge of neural networks and AI concepts. Sequential behavior is essential to intelligence and a fundamental part of human activities ranging from reasoning to language, and from everyday skills to complex problem solving. Sequence learning is an important component of learning in many task domains --- planning, reasoning, robotics, natural language processing, speech recognition, adaptive control, time series prediction, and so on. Naturally, there are many different approaches towards sequence learning. These approaches deal with somewhat differently formulated sequential learning problems, and/or different aspects of sequence learning. This book will provide an overall framework for this field of study. Table of Contents Introduction to Sequence Learning, Ron Sun Part 1: Sequence Clustering and Learning with Markov Models - Sequence Learning via Bayesian Clustering by Dynamics, Sebastiani et al - Using Dynamic Time Warping to Bootstrap HMM-Based Clustering of Time Series, Oates et al Part 2: Sequence Prediction and Recognition with Neural Networks - Anticipation Model for Sequential Learning of Complex Sequences, Wang - Bidirectional Dynamics for Protein Secondary Structure Prediction, Baldi et al - Time in Connectionist Models, Chappelier et al - On the Need for a Neural Abstract Machine, Sona et al Part 3: Sequence Discovery with Symbolic Methods - Sequence Mining in Categorical Domains: Algorithms and Applications, Zaki - Sequence Learning in the ACT-R Cognitive Architecture: Empirical Anal- ysis of a Hybrid Model, Lebiere et al Part 4: Sequential Decision Making - Sequential Decision Making Based on Direct Search, Schmidhuber - Automatic Segmentation of Sequences through Hierarchical Reinforcement Learning, Sun et al - Hidden-Mode Markov Decision Processes for Nonstationary Sequential De- cision Making, Choi et al - Pricing in Agent Economies Using Neural Networks and Multi-agent Q- learning, Tesauro Part 5: Biologically Inspired Sequence Learning Models - Multiple Forward Model Architecture for Sequence Processing, Bapi et al - Integration of Biologically Inspired Temporal Mechanisms into a Cortical Framework for Sequence Processing, Frezza-Buet et al - Attentive Learning of Sequential Handwriting Movements: A Neural Net- work Model, Grossberg et al ------------------------------ End of ML-LIST Digest Vol 13, No. 1 ***********************************