Received: from ics.uci.edu by Paris.ics.uci.edu id aa10708; 22 Sep 92 17:54 PDT Received: from ics.uci.edu by q2.ics.uci.edu id aa20231; 22 Sep 92 17:54 PDT To: ML-LIST: ; Subject: Machine Learning List: Vol. 4 No. 19 Reply-to: ml@ics.uci.edu Date: Tue, 22 Sep 92 17:41:45 -0700 From: Michael Pazzani Message-ID: <9209221754.aa20231@q2.ics.uci.edu> Machine Learning List: Vol. 4 No. 19 Tuesday, Sept 22, 1992 Contents: Multistrategy Learning Systems European Post-Doc grants UCI Machine Learning Repository ECML-93 Call for Proposals to Organize 1994 ML Conference The Machine Learning List is moderated. Contributions should be relevant to the scientific study of machine learning. Mail contributions to ml@ics.uci.edu. Mail requests to be added or deleted to ml-request@ics.uci.edu. Back issues may be FTP'd from ics.uci.edu in pub/ml-list/V/ or N.Z where X and N are the volume and number of the issue; ID: anonymous PASSWORD: ---------------------------------------------------------------------- Subject: multistrategy learning systems Date: Wed, 16 Sep 92 13:53:12 N From: morik@kilo.informatik.uni-dortmund.de The following is a reply to the request of Dick Jackson on multistrategy learning systems. The European Community funds a project, the Machine Learning Toolbox (MLT), where several learning algorithms and a statistical package are made available in an environment with a common look and feel. Moreover, the Common Knowledge Representation Language CKRL is the data format which can be read by all the integrated systems. It is translated into the internal representation so that the system uses its appropriate representation form for learning. The learning result is then re-translated into CKRL. So, the user of the MLT represents the data in CKRL and calls various learning or statistical algorithms. The user is supported in choosing the best suited algorithm by a CONSULTANT. The MLT includes statistical algorithms (provided by INRIA, France), a knowledge-intensive clustering system using a restricted first order logic, KBG (Univ. Paris-Sud,F), a variant of ID3 called NewID (Turing Institute, UK), a variant of AQ and ID called CN2 (Turing Institute, UK), a learning apprentice in a restricted first-order logic called APT (ISoft and Univ. Paris-Sud, F), an algorithm for similarity based discrimination called LASH (British Aerospace, UK), a multistrategy learning system called MOBAL which integrates model-based learning in first-order logic and constructing a lattice of constant terms into a knowledge acquisition system (GMD, Germany). The project coordinator is Marc Uszynski, Alcatel Alsthom Recherche, e-mail: mlt@aar.alcatel-alsthom.fr ------------------------------ Subject: European Post-Doc grants Date: Wed, 16 Sep 92 14:52:03 N From: morik@kilo.informatik.uni-dortmund.de Post-doc grants in Europe -- University of Dortmund -- Germany ---- The Commission of the European Communities, Directoriate general 12 in its program 'Human Capital and Mobility' funds young European researchers, especially at post-doctoral level, but in some cases also before the PhD thesis, for them working two years at another European country. One of the host institutions is the University of Dortmund, Germany. The university of Dortmund offers scientists the opportunity of being trained in a young and excellent team. Our research is focused on exploiting machine learning in applications from robotics to databases. The different projects are: behavioral learning - applying machine learning to robots, inductive logic programming, learning in humans and machines, applying machine learning to databases. Young European (but not German) scientists being interested in working with us should write an informal application to Prof. Dr. Katharina Morik, morik@kimo.informatik.uni-dortmund.de Univ. Dortmund, Fb 4, LS 8, POBox 500 500, D-4600 Dortmund 50 ------------------------------ Subject: UCI Machine Learning Repository Date: Mon, 21 Sep 92 20:18:29 -0700 From: "Patrick M. Murphy" The following is a list of recently added (at least since the last update posting to ML-List) and recently documented (moved from the undocumented/ directory) databases and domain theories. Any comments or donations would be greatly appreciated. Patrick M. Murphy (Site Librarian) David W. Aha (Off-Site Assistant) 1. Standard Attribute version of Audiology Database (donated by Ross Quinlan) The original audiology database has a non-standard attibute description. This database is a standardized version where the instances are described in terms of 69 (mostly binary) attributes according to a well described set of rules. 2. Wisconsin Breast Cancer Database (donated by Olvi Mangasarian) A two class breast cancer database with 699 instances of 9 integer-valued attributes, with a small number of missing values. One recent usage is Zhang, J. MLC 92. 3. Japanese Credit Screening Database (donated by Chiharu Sano) A credit application screening database from Japan. Includes a domain theory generated via discussions with Japanese domain experts. 125 instances. 4. Credit Card Application Approval Database (donated by Ross Quinlan) Contains 690 instances of 15 attributes, some with missing values. Interesting because of its good mix of attributes -- continuous, nominal with small numbers of values, and nominal with larger numbers of values. 5. Horse Colic Database (Mary McLeish & Matt Cecile) 368 instances with 28 attributes (continuous, discrete, and nominal), 30% missing values. Multiple reasonable class attributes. Very well documented attribute descriptions. 6. Image Segmentation Database (Carla Brodley: UMass) The instances were drawn randomly from a database of 7 outdoor images (brickface, sky, foliage, cement, window, path, grass). The images were handsegmented to create a classification for every pixel. Each instance represents a 3x3 region. 2310 instances of 19 continuous attributes. 7. Johns Hopkins University Ionosphere Database (V. Sigillito) This radar data was collected by a system in Goose Bay, Labrador. This system consists of a phased array of 16 high-frequency antennas with a total transmitted power on the order of 6.4 kilowatts. The targets were free electrons in the ionosphere. The classes: "Good" radar returns are those showing evidence of some type of structure in the ionosphere, "Bad" returns are those that do not (these signals pass through the ionosphere). 351 instances of 34 continuous attributes. No missing values. 8. Lung Cancer Database (Donated by Stefan Aeberhard) The data described 3 types of pathological lung cancers. The Authors give no information on the individual variables nor on where the data was originally used. This data was used by Hong and Young to illustrate the power of the optimal discriminant plane even in ill-posed settings. Contains 32 instances of 57 Attributes. 9. Primate Splice-Junction Gene Sequences (DNA) with associated Imperfect Domain Theory (donated by G. Towell, M. Noordewier, and J. Shavlik) Splice junctions are points on a DNA sequence at which `superfluous' DNA is removed during the process of protein creation in higher organisms. The problem posed in this dataset is to recognize, given a sequence of DNA, the boundaries between exons (the parts of the DNA sequence retained after splicing) and introns (the parts of the DNA sequence that are spliced out). This problem consists of two subtasks: recognizing exon/intron boundaries (referred to as EI sites), and recognizing intron/exon boundaries (IE sites). (In the biological community, IE borders are referred to a ``acceptors'' while EI borders are referred to as ``donors''.) 3190 instances of 62 attributes. 10. Pima Indians Diabetes Database (donated by Vincent Sigillito) The diagnostic, binary-valued variable investigated is whether the patient shows signs of diabetes according to World Health Organization criteria (i.e., if the 2 hour post-load plasma glucose was at least 200 mg/dl at any survey examination or if found during routine medical care). The population lives near Phoenix, Arizona, USA. Several constraints were placed on the selection of these instances from a larger database. In particular, all patients are females at least 21 years old of Pima Indian heritage. 768 instances of 8 numeric attributes. 11. Quadraped Animals Data Generator (donated by John Gennari) A data generator of structured instances representing quadruped animals as used by Gennari, Langley, and Fisher (1989) to evaluate the CLASSIT unsupervised learning algorithm. Instances have 8 components: neck, four legs, torso, head, and tail. Each component is represented as a simplified/generalized cylinder (i.e., inspired by David Marr's work in "Vision: A Computational Investigation Into the Human Representation and Processing of Visual Information", published by Freeman in 1982). Each cylinder is itself described by 9 attributes: location x 3, axis x 3, height, radius, and texture. This code generates instances in one of four classes: dogs, cats, horses, and giraffes. The program generates instances by selecting a class according to a distribution determined by function rand4(). Each class has a prototype; the prototype of the selected class is perturbed according to a distribution described in the code for the four classes (i.e., parameterized means with Guassian distributions are used to represent prototypes and perturbation distributions, where the means are used to distinguish the four classes). 12. Solar Flare Database (donated by Gary Bradshaw) The database contains 3 potential classes, one for the number of times a certain type of solar flare occured in a 24 hour period. Each instance represents captured features for 1 active region on the sun. 1389 instances of 13 attributes; no missing values. 13. Tic-Tac-Toe Endgame Database (donated by David Aha) This database encodes the complete set of possible board configurations at the end of tic-tac-toe games, where "x" is assumed to have played first. The target concept is "win for x" (i.e., true when "x" has one of 8 possible ways to create a "three-in-a-row"). 958 instances, all attributes can take on 1 of 3 possible values. 14. Thyroid Gland Database (donated by Stefan Aeberhard) Five lab. tests are used to try to predict whether a patient's thyroid belonged to the class euthyroidism, hypothyroidism or hyperthyroidism. The diagnosis was based on a complete medical record, including anamnesis, scan etc. 215 instances, no missing values. 15. Wine Recognition Database (donated by Stefan Aeberhard) These data are the results of a chemical analysis of wines grown in the same region in Italy but derived from three different cultivars. The analysis determined the quantities of 13 constituents found in each of the three types of wines. 178 instances of 13 continuously valued attributes. 16. A Newer Version of the Othello Domain Theory (donated by Tom Fawcett) This theory is used in research to generate features for an inductive learning system. Coded in Prolog. This version is similar to the older version except that it has been cleaned up and has had some meta declarations added. ------------------------------ Date: Thu, 17 Sep 1992 17:16:59 +0100 From: Pavel Brazdil Subject: ECML-93 ECML-93 European Conference on Machine Learning ======================================= 5-7 April 1993 Vienna, Austria Second Announcement and Call for Papers (Reminder) General Information: ECML-93 will continue with the tradition of earlier EWSL's (European Working Session on Learning) and provide a platform for presenting the latest results in the area of machine learning. Although ECML-93 is the first conference under this name, it can be considered as the sixth meeting of this kind in Europe. Programme: The scientific programme will include presentation of selected papers and several invited talks. One will be presented by Ross Quinlan from the University of Sydney. Another will be presented by Derek Sleeman who will cover some European research projects in the area of ML and their significance. Stephen Muggleton will deliver a lecture on Inductive Logic Programming (ILP). The programme will be complemented by the following workshops which will take place immediately after the main conference (8 April): W1: Integrated Learning Architectures W2: Foundations of Evolutionary Computation W3: ML Techniques and Text Analysis W4: Learning Robots W5: Industrial Applications of ML Submission of Papers: Submissions are invited on original research covering all aspects of machine learning including, but not limited to: learning system architectures multi-strategy learning inductive & deductive methods inductive logic programming abduction automated discovery representational change learning in problem solving reinforcement learning learning by analogy case-based learning unsupervised learning neural network learning genetic approaches theory of learnability evaluation of learning methods applications of ML Long papers should be limited to 18 pages. Short papers describing the work in progress should be limited to 10 pages. Submissions should be made in four copies to the Programme Chairman. Important Dates: Submission deadline: 16 October 1992 <=== reminder Notification of acceptance: 4 December 1992 Camera ready copy: 15 January 1993 Programme Chairman: Pavel Brazdil Tel.: (+351) 2 600 1672 Ext. 106 LIACC, Rua Campo Alegre 823 Fax: (+351) 2 600 3654 4100 Porto, Portugal email: pbrazdil @ ncc.up.pt Programme Committee: F. Bergadano (Italy) I. Bratko (Slovenia) P. Brazdil (Portugal) L. de Raedt (Belgium) J. G. Ganascia (France) K. de Jong (USA) A. Kakas (Cyprus) Y. Kodratoff (France) N. Lavrac (Slovenia) R.L. de Mantaras (Spain) K. Morik (Germany) I. Mozetic (Austria) S. Muggleton (UK) L. Saitta (Italy) D. Sleeman (UK) J. Shavlik (USA) M. Someren (Netherl.) W. Van de Velde (Belgium) R. Wirth (Germany) Local Arrangements: Igor Mozetic and Gerhard Widmer Austrian Research Institute for AI and Dept. of Medical Cybernetics and AI Schottengasse 3 A - 1010 Vienna, Austria Email: ecml @ ai.univie.ac.at Tel.: (+43) 1 533 6112 or (+43) 1 535 32810 Fax: (+43) 1 532 0652 ECML-93 Workshops ==================== 8 April 1993, Viena, Austria General Information The European Conference on Machine Learning (ECML-93) that will take place in Viena, Austria 5-7 April 1993 will be supplemented by several workshops immediately after the main conference, that is, 8 April 1993. The programme of these workshops is entirely in the hands of the workshop organisers. The organisers of the main conference venue will however provide the necessary logistic support for each workshop. This will including distribution of announcements, preparation of proceedings (in the form of handouts), and necessary space. Details about the individual ECML-93 workshops follow. This document contains provisional information about the workshops. Conditions of Attendance All participants wishing to assist an ECML-93 workshop must register for the main conference, and in addition satisfy the criteria spelled out by each workshop organizer (see individual workshop announcements for details). No extra fee will be charged for the attendance. Deadline for Submissions Unless the individual workshop announcements state otherwise, the deadline for submissions to workshops is 31 January 1993. All submissions should be sent directly to the chair of the workshop. W1: Integrated Learning Architectures ======================================= Main Themes: How do learning and problem solving constrain or support each other? What is the goal of learning? How the learning goals are generated and selected? What knowledge does a learning method require in order to be able to learn? Is that knowledge explicit in a (meta-)model? How the results of learning are integrated into the overall architecture? What are the implications of an integrated architecture for issues such as bias, utility and knowledge revision? Chair: Enric Plaza CEAB - CSIC, Cam de Santa Barbara, 17300 Blanes, Catalunya, Spain email: plaza@ceab.es, Fax: (+34) 7233 7806 Programme Committee: Agnnar Aadmondt (Norway), Nada Lavrac (Slovenia), Ashwin Ram (USA), Walter Van de Velde (Belgium), Maarten van Someren (Netherlands). W2: Foundations of Evolutionary Computation ============================================== Main Themes: Theories of evolutionary computation: The theories should contrast and compare different evolutionary computation approaches, such as genetic algorithms, evolutions strategies, evolutionary programming, and genetic programming. Comparison of different computation approaches on machine learning tasks. These may be theoretical and experimental. Chair: William M. Spears Navy Center for Applied Research in AI Naval Research Laboratory, Code 5510, Washington D.C., 20375-5320 USA email: spears@aic.nrl.navy.mil Co-chairs: Kenneth A. De Jong (USA), Gilles Venturini (France) W3: ML Techniques and Text Analysis ===================================== Main Themes: Syntactic learning. Semantic learning. New/existing ML techniques geared to NL. Statistical techniques applied to NL. Lexical disambiguation. Linguistic pattern matching. Lexical acquisition. Automatic analysis of bi-lingual corpora. Information theoretic results and measures. Chair: Peter Adriaans Syllogic B.V., Postbus 26,3990 DA Houten, The Netherlands email: pieter@syllogic.nl, Fax: (+31) 3403 51095 Programme Committee and/or Participants: Lars Asker, Bob Berwick (USA), Walter Daelemans (Netherlands), Pat Langley (USA), Mitch Markus (USA), Stan Matwin (Canada), Laurent Miclet (France), David Powers (Germany), Remco Scha (Netherlands), Jeff Siskind (USA), Won-Wun Soo (Taiwan), Larry Reekeer, Gerry Wolff (UK), Many Reiner (USA), Wendy Lehnert (USA). W4: Learning Robots =========================== Main Themes: Applications of ML in robotics including the following sub-themes: Integration of symbolic and subsymbolic learning. Learning in noisy environments and structured domains. Learning heuristics in planning tasks. Real world applications. Chair: Attilio Giordana Univ. of Torino, Dip. di Informatica, Corso Swizzera 185, 10149 Torino, Italy email: attilio@di.unito.it, Fax: (+39) 11 759 603 Programme Committe: R.Dillmann (Germany), A.Steiger-Gar o (Portugal), Y.Kodratoff (France), J.Millan (Spain), K.Morik (Germany), M.Golumbic (Israel) W5: Industrial Applications of ML ========================================= Chair: Yves Kodratoff LRI, Bat 490, Univ. Paris-Sud, 91405 Orsay, France email: yk@sun8.lri.fr Fax: (+33) 1 6941 6586 ------------------------------ Date: Fri, 11 Sep 92 14:40:19 PDT From: Tom Dietterich Subject: Call for Proposals to Organize 1994 ML Conference In June 1994, the Eleventh International Machine Learning Conference will be held. The purpose of this call is to invite groups interested in organizing and hosting the conference to submit proposals. The group selected to run the conference will be given full authority and responsibility for producing the conference. Proposals should address the following issues: 1. Organization and Format. In previous years, the format of the conference has alternated annually between a single plenary session (1988, 1990) and a set of parallel workshop sessions (1989, 1991). However, in both 1992 and 1993, the format involved 3 days of plenary sessions followed by one day of specialized workshops. A poster session has also been held at each conference. Please indicate what format you would like to have and how you would arrange the schedule to suit the format. In the past, the conference has been organized by an Organizing Committee whose membership included organizers of past conferences and other senior researchers. Review of papers has been conducted by a Program Committee selected by the organizers with the advice of the Organizing Committee. Please indicate what organization you would employ. 2. Locale Parameters. - Accessibility. Is it easy and inexpensive for people (especially graduate students) to travel to the conference site? (Compute mean airfares from Europe and North America.) - Meeting Rooms, AV Equiment, etc. What are the physical facilities like? - Meals and Lodging. Is there low-cost, quality housing available for attendees (especially graduate students)? How far from the meeting rooms? Where will attendees eat? - Demo facilities. Will there be computing equipment and space available to support demos? 3. Local Machine Learning Community. Is there a local ML group/community that can help with organization and funding. 4. Organizational and Financial Support. Can the host institution(s) provide support for registration and financial management (e.g., credit card payments, accounting, etc.). How will the conference be funded? Provide a draft budget covering expenses, expected registration fee schedule, and sources of financial support. The host institution must agree to forward any unused funds to the host of the 1995 conference. In previous years, funding has been obtained from federal granting agencies, corporations, and universities. Proposals should be sent before October 15, 1992 to Tom Dietterich Department of Computer Science 303 Dearborn Hall Oregon State University Corvallis, OR 97331-3202 As in previous years, the choice of organizers will be made by the Editorial Board of the Machine Learning Journal. ------------------------------ End of ML-LIST (Digest format) ****************************************