neuron-request@HPLMS2.HPL.HP.COM ("Neuron-Digest Moderator Peter Marvit") (08/18/90)
Neuron Digest Friday, 17 Aug 1990 Volume 6 : Issue 49 Today's Topics: Re: universe and intelligence Really smart systems Re: Really smart systems Re: Really smart systems tasks for reinforcement learning the dumb universe Help for RTRL? Re: Help for RTRL? PYGMALION Overview NN-definition Language Re: NN-definition Language Last Call for Papers for AGARD Conference Send submissions, questions, address maintenance and requests for old issues to "neuron-request@hplabs.hp.com" or "{any backbone,uunet}!hplabs!neuron-request" Use "ftp" to get old issues from hplpm.hpl.hp.com (15.255.176.205). ------------------------------------------------------------ Subject: Re: universe and intelligence From: arti6!chen@relay.EU.net (Chung-Chih Chen) Date: Wed, 15 Aug 90 12:40:22 +0100 Concerning the reply from Douglas Danforth: >Intelligence is in the mind of the intelligent AND why do you consider >intelligence significant? What I really mean in my last message can be explained more clearly as follows: According to the inflationary model of the universe (see, for example, the article by A. Guth & P. Steinhardt in the book edited by P. Davies), the universe could have been created from virtually nothing (so the universe may be the ultimate free lunch!). If the seemly very complex behaviors of the current universe came from nothing, then can we imagine that we may build a universal model of our brain (or neural network) which will produce intelligent behaviors from nothing? Nolfi et al. showed that complex and apparently purposeful behavior can arise from random variation in networks. In some way they have shown that intelligence can come from nothing (or evolution). I hope this will clarify my idea. %E P. Davies %B The New Physics %I Cambridge University Press %D 1989 %A S. Nolfi %A J. L. Elman %A D. Parisi %T Learning and Evolution in Neural Networks %J CRL Technical Report 9019 %I Center for Research in Language, University of California San Diego %D July 1990 ------------------------------ Subject: Really smart systems From: kingsley@hpwrce.HP.COM (Kingsley Morse) Organization: Ye Olde Salt Mines Date: 14 Aug 90 21:05:25 +0000 If we aspire to making truly intelligent machines, our algorithms must scale up well to large training sets. In other words, the computational complexity of our algorithms must accomodate many training patterns and many dimensions. You may have heard of the "curse of dimensionality". If the computational complexity of our algorithms grows slowly, then we can train with more patterns and dimensions to get a smarter system. On the other hand, if the computation required grows explosively as the number of training patterns or dimensions are increased, then even astoundingly fast hardware won't help. So, our challange is to find algorithms which scale up well to large problems, so we can make really smart systems. Listen........... I'll post some computational complexity figures for some common algorithms. If you add to (correct?) these, please do. The terminology that I propose is that linear computational complexity is better than polynomial which is better than exponential. Furthermore, the algorithms' scalability must be measured with respect to learning, recall, patterns, and dimensions. Algorithm Learning Recall patterns dimensions patterns dimensions ------------------------------------------------------------ Backprop polynomial ? independent linear Nearest neighbor linear linear linear linear Cart nlogn exponential independent ? This leads me to believe that nearest neighbor algorithms work better for learning a lot because they can be trained faster. Any contributions to this list are welcome. ------------------------------ Subject: Re: Really smart systems From: mcsun!ukc!reading!minster!russell@uunet.uu.net Organization: Department of Computer Science, University of York, England Date: 17 Aug 90 11:21:56 +0000 Well, if you want tuppeny-worth's thrown in, here's mine. Backprop has NO guarenteed convergence, therefore to quote computational complexity is misleading. It may be possible to give an "average case" complexity figure, but this is so dependent on the initial weight settings as to be not too much use. Tesauro and Janssens report empirical results between learning time and predicate order (q) of patterns. The net has q inputs, 2q nodes in the first layer, fully connected, and 1 output node. The task set is the parity function on the q bits. Leaning times in b-p scale as approx. 4^q. The task set is of size 2^q so the training time is about 2^q. Conclusion is that empirically learning time is of exponential order. The perceptron scales as exponential in input size (Hampson and Volper 1986). (See Neural Network Design and the Complexity of Learning by Judd for more.....I aint read it all yet, but it's interesting so far.....) I confess to not understanding quite what "Cart" means in the above table - I can make an educated guess, however..... To add another algorithm to the list, the ADAM system, based on the Willshaw net (i.e. a distributed associative memory) (Austin 1987) has complexity as follows: Algorithm Learning Recall patterns dimensions patterns dimensions ------------------------------------------------------------ ADAM linear linear(quadratic) indep. linear(nlogn) brackets refer to abnormal, but permissible, parameterisation. (Beale 1990). Note that recall may take longer than teaching - but also note that the realm of use of ADAM means that the multiplicative constants in front of the order terms are extremely small. Russell. ------------------------------ Subject: Re: Really smart systems From: usenet@tut.cis.ohio-state.edu (usenet news poster) Organization: National Library of Medicine, Bethesda, Md. Date: 17 Aug 90 21:19:11 +0000 >From a separate communiction with km: CART means "Classification and Regression Tree". It is similar to ID3, and here's how it works. The training vectors are used to "grow" a decision tree. The decision tree can be used catagorize new inputs. OK, let me add a couple more cents to the pot. Classifications need to consider both the computational complexity and the storage requirements. Algorithms like nearest neighbor, CART? and ADAM? require that the complete training set be stored for recall while a perceptron needs storage of the order #inputs+#outputs, and a single hidden layer fully connected neural net needs (#inputs+#output)*#hidden_nodes. A second consideration is will the algorithm generalize? Ie., will it make use of information from more than one input pattern to formulate an output. So at the risk of grossly misquoting and being flamed horribly let me reorder the classification in terms of Np=#patterns, Ni=#inputs, No=#outputs, and Nh=#hidden nodes: Algorithm Learning time Recall time Storage Generalizes ----------------------------------------------------------------------------- Perceptron No*(x^Ni) No*Ni No*Ni limited Neural Net (1 hidden layer) Nh*(Ni+No) Nh*(Ni+No) yes Backprop x^(Nh*(Ni+No)) ? Conjugate gradient (Nh*(Ni+No))^3 - locally Monte Carlo x^(Nh*(Ni+No)) ? Nearest neighbor (Ni+No)*Np Ni*Np Np*(Ni+No) no Class & Reg. Tree Ni*(Np log Np) Ni*log Np Np*(Ni+No) no (CART) + (Ni+No)*Np ? + Ni*log Np Adaptive Memory Np? ? Np*(Ni+No)? no? (ADAM) David States ------------------------------ Subject: tasks for reinforcement learning From: finton@ai.cs.wisc.edu (David J. Finton) Organization: U of Wisconsin CS Dept Date: 15 Aug 90 17:28:08 +0000 I'm looking for good demonstration tasks for my reinforcement-learning algorithm: (1) Are there any good examples of real-world tasks which require reinforcement learning -- where supervised techniques such as back-prop would be unsuitable? (2) Are there any studies which compare performance of reinforcement learning systems with standard techniques (eg, back-prop, ID3) on such tasks? (3) Are there available standard data sets for such tasks? (4) Are there studies comparing reinforcement learning on with back-prop on standard back-prop tasks? David Finton ------------------------------ Subject: the dumb universe From: Stephen Smoliar <smoliar@vaxa.isi.edu> Date: Wed, 15 Aug 90 17:38:37 -0700 There is no reason to assume that intelligence on the part of the universe is prerequisite to it giving rise to intelligent systems. That is the whole point of THE BLIND WATCHMAKER. It is also the point of THE SOCIETY OF MIND. Minsky's argument is that you can build an intelligent system from lots of little components, each of which is far to simple to be, in itself, intelligent. Neurons are an example of such little components. ------------------------------ Subject: Help for RTRL? From: coms2146@waikato.ac.nz (Alistair Veitch, University of Waikato, New Zealand) Organization: University of Waikato, Hamilton, New Zealand Date: 16 Aug 90 03:54:44 +0000 Has anybody out there worked with Williams and Zipsers "Real-time recurrent learning algorithm"? [Connection Science, Vol 1, No 1]. We are currently trying to implement this algorithm, but have run into some problems. We've got it to run succesfully on the various XOR problems described, the "ab" problem (recognise the first "b" after an "a") and the oscillation problems. What we can't seem to achieve is success for the Turing machine problem. As this is perhaps the major result of the paper, it seems important to duplicate it to reassure ourselves that everything is correct. Has anyone else had success/failure with this problem? If success, would it be possible to post your source? (We think we've got it right, but...) Alistair Veitch Phone: +64 71 562889 ext. 8768 Internet: coms2146@waikato.ac.nz +64 71 562388 (home) SNAIL: Computer Science Dept, University of Waikato, Hamilton, New Zealand ------------------------------ Subject: Re: Help for RTRL? From: ins_atge@jhunix.HCF.JHU.EDU (Thomas G Edwards) Organization: The Johns Hopkins University - HCF Date: 16 Aug 90 19:38:09 +0000 In article <1243.26cac1c4@waikato.ac.nz> coms2146@waikato.ac.nz (Alistair Veitch, University of Waikato, New Zealand) writes: >Has anybody out there worked with Williams and Zipsers "Real-time recurrent >learning algorithm"? [Connection Science, Vol 1, No 1]. I haven't actually implemented this algorithm, but I have heard that it is important to use the "Teacher Forcing" method they discuss to learn difficult problems. You might also want to look at J. Schmidhuber, "Making the World Differentiable: On using supervised learning fully-recurrent networks for dynamic reinforcement learning and planning in non-stationary environments", FKI Report 125-90, Technische Univeritat Munchen, 1990. A pole-balancer is trained by reinforcement learning (i.e. apply pain when the pole is dropped). And to explain why gradient-descent methods will probably not give you reasonable temporal learning see J. Schmidhuber, "Towards compositional learning with dynamic neural networks", FKI Report 129-90, TUM, April 1990. He explains that gradient-descent-only methods must take into account training learned during all past time steps when dealing with a new problem. For "toy" temporal learning problems, this is not a big impediment. For "serious" temporal learning problems, dynamic neural systems must develop methods of breaking goals down into subgoals, most of which have already been learned, some of which need to be developed by gradient-descent. In this way, only small problems are trained by gradient-descent, and they are used by the system combinatorially to allow the network-of-networks to solve real problems by "divide-and-conquer" methods. The research is very fresh into this area, and I think in about a year there will be a move away from naive implementations of gradient-descent learning in both stationary and temporal learning and a move towards connectionist compositional learning (Cascade-Correlation is a simple example of this). -Thomas Edwards ------------------------------ Subject: PYGMALION Overview From: M.Azema@cs.ucl.ac.uk Date: Fri, 17 Aug 90 10:01:52 +0100 In response to requests about the PYGMALION environment, (with some delays) here is an overview: PYGMALION Overview: ------------------- The ESPRIT II PYGMALION project is intended to provide a focus for neural computing research within the European Community. PYGMALION aims to promote the application of neural networks by European industry, and to develop European "standard" computational tools for programming and simulation of neural networks. The design philosophy of the PYGMALION neural programming environment is twofold. Firstly, to provide an "open" programming environment - a rudimentary "platform" - that can be easily extended and interfaced to other tools. For this reason the core of the environment is X-windows, C and C++; running on a colour workstation. Secondly, to provide "portable" neural network applications, so that trained and partially trained networks can be easily moved from machine to machine. For this reason the (partially) trained neural network applications are specified in a subset of C; essentially a C data structure. The environment comprises 5 major parts: Graphic Monitor, the graphical software environment for controlling the execution and monitoring of a neural network application simulation. This includes a simulation command language for setting up a simulation, monitoring its execution, interactively changing values, and saving a trained network. Algorithm Library, the parameterised library of common neural networks, written in the high level language and providing the user with a number of validated modules for constructing applications. High Level Language N, the object-oriented programming language for defining, in conjunction with the algorithm library, a neural network algorithm and application, by describing the network topology and its dynamics. Intermediate Level Language nC-code, the low level machine independent network specification language for representing the partially trained or trained neural network applications, a format analogous to P-code for PASCAL systems. Compilers to the target UNIX-based workstations and parallel Transputer-based machines. Availability of the software: ----------------------------- A preliminary version of the PYGMALION software is now available (free of charge). If you would like more information please contact : Mike Hewetson Department of Computer Science University College London Gower Street London WC1E 6BT Voice: +44 (0) 71 387 7050 ext 3708 Fax: +44 (0) 71 387 1397 Email: M.Hewetson@cs.ucl.ac.uk ------------------------------ Subject: NN-definition Language From: ethz!neptune!brain!thalmann@uunet.uu.net (Laura Thalmann) Organization: Department for Informatik,Universitat Zurich-Irchel Date: 17 Aug 90 11:23:54 +0000 Hi neural-experts, This is a presentation of (yet another) neural network implementation, a NN-definition language: Condela means CONnection DEfinition LAnguage and it is a high level programming language, specifically designed for the development and modeling of neural network applications. It is a procedural and general purpose language, that allows parallelism via the concept of selections, i.e. groups of units or connections to which actions can be applied. Units and connections can be created dynamically at any point in the program flow. The parallelism expressible in Condela-3 is independent of the underlying Hardware. Condela-3 is easy to teach, as it has few language constructs, yet allows the expression of arbitrary network topologies and learning paradigms due to its powerful statements and its two levels of abstraction. It is easily portable to other operating systems, its open design allows simple interfacing to existing applications. The following sample program demonstrates the classical XOR learning problem using error back propagation. 1 TOPOLOGY 2 xor = LAYER input OF FIELD[2]; END; 3 LAYER hidden OF FIELD[2]; END; 4 LAYER output OF FIELD[1]; END; 5 VAR p : NETWORK OF xor; 6 7 PROCEDURE main(); 8 VAR output_vec, input_vec : VECTOR; 9 input_layer, hidden_layer, output_layer : USEL; 10 11 BEGIN 12 CREATE p; 13 input_layer := { p.input[0..1] }; 14 hidden_layer := { p.hidden[0..1] }; 15 output_layer := { p.output[0] }; 16 CONNECT input_layer TO hidden_layer INIT random(); 17 CONNECT hidden_layer TO output_layer INIT random(); 18 LOOP 1000000 TIMES 19 get_input(input_vec, output_vec); 20 input_layer : out := input_vec; 21 APPLY feed_forward() TO hidden_layer; 22 APPLY feed_forward() TO output_layer; 23 APPLY back_propagate_out(output_vec) TO output_layer; 24 APPLY back_propagate_hid() TO hidden_layer; 25 END; 26 END; It has a 2 layered implementation that allows the "abstract" definition of a neural network topology and behavior and a "concrete" implementation in C. This compiler (implemented with lex and yacc) translates the Condela-source to C and therefore allows simple interfacing to other existing neural network simulation systems. I appreciate any comments. -Nick. ,----------------------------------------------------, | Nikolaus Almassy almassy@ifi.unizh.ch / | University of Zurich-Irchel Tel:+41-1-257 43 15 / | Department of Informatik Fax:+41-1-257 43 43 / | Winterthurerstr. 190 CH-8057 SWITZERLAND / `-----------------------------------------------' ------------------------------ Subject: Re: NN-definition Language From: van-bc!ubc-cs!kiwi!snider@ucbvax.Berkeley.EDU (Duane Snider) Organization: Microtel Pacific Research Ltd., Burnaby, B.C., Canada Date: 17 Aug 90 17:45:19 +0000 > [[CONDELA]] It is a procedural and ^^^^^^^^^^ >general purpose language, that allows parallelism via the concept of ^^^^^^^^^^^^^^^^^^^^^^^^ It appears CONDELA isn't doing anything more than a programming language like C++ could handle. Are you sure that another language is necessary in this field, yet? Duane Snider snider@mpr.ca ------------------------------ Subject: Last Call for Papers for AGARD Conference From: nelsonde%avlab.dnet@wrdc.af.mil Date: Mon, 13 Aug 90 10:10:04 -0400 Subject: Last Call for Papers for AGARD Conference We are extending the deadline for the abstracts for the papers to be presented at the AGARD conference until 21 September 1990. In case you have lost the Call for Papers, it is again attached to this message. Your consideration is greatly appreciated. --Dale AGARD ADVISORY GROUP FOR AEROSPACE RESEARCH AND DEVELOPMENT 7 RUE ANCELLE - 92200 NEUILLY-SUR-SEINE - FRANCE TELEPHONE: (1)47 38 5765 TELEX: 610176 AGARD TELEFAX: (1)47 38 57 99 AVP/46 2 APRIL 1990 CALL FOR PAPERS for the SPRING, 1991 AVIONICS PANEL SYMPOSIUM ON MACHINE INTELLIGENCE FOR AEROSPACE ELECTRONICS SYSTEMS to be held in LISBON, Portugal 13-16 May 1991 This meeting will be UNCLASSIFIED Abstracts must be received not later than 31 August 1990. Note: US & UK Authors must comply with National Clearance Procedures requirements for Abstracts and Papers. THEME MACHINE INTELLIGENCE FOR AEROSPACE ELECTRONICS SYSTEMS A large amount of research is being conducted to develop and apply Machine Intelligence (MI) technology to aerospace applications. Machine Intelligence research covers the technical areas under the headings of Artificial Intelligence, Expert Systems, Knowledge Representation, Neural Networks and Machine Learning. This list is not all inclusive. It has been suggested that this research will dramatically alter the design of aerospace electronics systems because MI technology enables automatic or semi-automatic operation and control. Some of the application areas where MI is being considered inlcude sensor cueing, data and information fusion, command/control/communications/intelligence, navigation and guidance, pilot aiding, spacecraft and launch operations, and logistics support for aerospace electronics. For many routine jobs, it appears that MI systems would provide screened and processed ata as well as recommended courses of action to human operators. MI technology will enable electronics systems or subsystems which adapt or correct for errors and many of the paradigms have parallel implementation or use intelligent algorithms to increase the speed of response to near real time. With all of the interest in MI research and the desire to expedite transition of the technology, it is appropriate to organize a symposium to present the results of efforts applying MI technology to aerospace electronics applications. The symposium will focus on applications research and development to determine the types of MI paradigms which are best suited to the wide variety of aerospace electronics applications. The symposium will be organizaed into separate sessions for the various aerospace electronics application areas. It is tentatively proposed that the sessions be organized as follows: SESSION 1 - Offensive System Electronics (fire control systems, sensor cueing and control, signal/data/information fusion, machine vision, etc.) SESSION 2 - Defensive System electronics (electronic counter measures, radar warning receivers, countermeasure resource management, situation awareness, fusion, etc.) SESSION 3 - Command/Control/Communications/Intelligence - C3I (sensor control, signal/data/information fusion, etc.) SESSION 4 - Navigation System Electronics (data filtering, sensor cueing and control, etc.) SESSION 5 - Space Operations (launch and orbital) SESSION 6 - Logistic Systems to Support Aerospace Electronics (on and off-board systems, embedded training, diagnostics and prognostics, etc.) GENERAL INFORMATION This Meeting, supported by the Avionics Panel will be held in Lisbon, Portugal on 13-16 May 1991. It is expected that 30 to 40 papers will be presented. Each author will normally have 20 minutes for presentation and 10 minutes for questions and discussions. Equipment will be available for projection of viewgraph transparencies, 35 mm slides, and 16 mm films. The audience will include Members of the Avionics Panel and 150 to 200 invited experts from the NATO nations. Attendance at AGARD Meetings is by invitation only from an AGARD National Delegate or Panel Member. Final manuscripts should be limited to no more than 16 pages including figures. Presentations at the meeting should be an extract of the final manuscript and not a reading of it. Complete instructions will be sent to authors of papers selected by the Technical Programme Committee. Authors submitting abstracts should insure that financial support for attendance at the meeting will be available. CLASSIFICATION This meeting will be UNCLASSIFIED LANGUAGES Papers may be written and presented either in English or French. Simultanewous interpretation will be provided between these two languages at all sessions. A copy of your prepared remarks (Oral Presentation) and visual aids should be provided to the AGARD staff at least one month prior to the meeting date. This procedure will ensure correct interpretation of your spoken words. ABSTRACTS Abstracts of papers offered for this Symposium are now invited and should conform with the following instructions: LENGTH: 200 to 500 words CONTENT: Scope of the Contribution & Relevance to the Meeting - Your abstract should fully represent your contribution SUMITTAL: To the Technical Programme committee by all authors (US authors must comply with Attachment 1) IDENTIFICATION: Author Information Form (Attachment 2) must be provided with you abstract CLASSIFICATION: Abstracts must be unclassified Your abstracts and Attachment 2 should be mailed in time to reach all members of the Technical Program Committee, and the Executive not later than 31 AUGUST 1990 (Note the exception for the US Authors). This date is important and must be met to ensure that your paper is considered. Abstracts should be submitted in the format shown on the reverse of this page. TITLE OF PAPER Name of Author Organization or Company Affiliation Address Name of Co-Author Organization or Company Affiliation Address The test of your ABSTRACT should start on this line. PUBLICATIONS The proceedings of this meeting will be published in a single volume Conference Proceedings. The Conference Proceedings will include the papers which are presented at the meeting, the questions/discussion following each presentation, and a Technical Evaluation Report of the meeting. It should be noted that AGARD reserves the right to print in the Conference Proceedings any paper or material presented at the Meeting. The Conference Proceedings will be sent to the printer on or about July 1990. NOTE: Authors that fail to provide the required Camera-Ready manuscript by this date may not be published. QUESTIONS concerning the technical programme should be addressed to the Technical Programme Committee. Administrative questions should be sent directly to the Avionics Panel Executive. GENERAL SCHEDULE (Note: Exception for US Authors) EVENT DEADLINE SUBMIT AUTHOR INFORMATION FORM 31 AUG 90 SUBMIT ABSTRACT 31 AUG 90 PROGRAMME COMMITTEE SELECTION OF PAPERS 1 OCT 90 NOTIFICATION OF AUTHORS OCT 90 RETURN AUTHOR REPLY FORM TO AGARD IMMEDIATELY START PUBLICATION/PRESENTATION CLEARANCE PROCEDURE UPON NOTIFICATION AGARD INSTRUCTIONS WILL BE SENT TO CONTRIBUTORS OCT 90 MEETING ANNOUNCEMENT WILL BE PUBLISHED IN JAN 91 SUBMIT CAMERA-READY MANUSCRIPT AND PUBLICATION/ PRESENTATION CLEARANCE CERTIFICATE to arrive at AGARD by 15 MAR 91 SEND ORAL PRESENTATION AND COPIES OF VISUAL AIDS TO THE AVIONICS PANEL EXECUTIVE to arrive at AGARD by 19 APR 91 ALL PAPERS TO BE PRESENTED 13-16 MAY 91 TECHNICAL PROGRAMME COMMITTEE CHAIRMAN Dr Charles H. KRUEGER Jr Director, Systems Avionics Division Wright Research and Development Center (AFSC), ATTN: AAA Wright Patterson Air Force Base Dayton, OH 45433, USA Telephone: (513) 255-5218 Telefax: (513) 476-4020 Mr John J. BART Prof Dr A. Nejat INCE Technical Director, Directorate Burumcuk sokak 7/10 of Reliability & Compatibility P.K. 8 Rome Air Development Center (AFSC) 06752 MALTEPE, ANKARA GRIFFISS AFB, NY 13441 Turkey USA Mr J.M. BRICE Mr Edward M. LASSITER Directeur Technique Vice President THOMSON TMS Space Flight Ops Program Group B.P. 123 P.O. Box 92957 38521 SAINT EGREVE CEDEX LOS ANGELES, CA 90009-2957 France USA Mr L.L. DOPPING-HEPENSTAL Eng. Jose M.B.G. MASCARENHAS Head of Systems Development C-924 BRITISH AEROSPACE PLC, C/O CINCIBERLANT HQ Military Aircraft Limited 2780 OEIRAS WARTON AERODROME Portugal PRESTEN, LANCS PR4 1AX United Kingdom Mr J. DOREY Mr Dale NELSON Directeur des Etudes & Syntheses Wright R & D Center O.N.E.R.A. ATTN: AAAT 29 Av. de la Division Leclerc Wright Patterson AFB 92320 CHATILLON CEDEX Dayton, OH 45433 France USA Mr David V. GAGGIN Ir. H.A.T. TIMMERS Director Head, Electronics Department U.S. Army Avionics R&D Activity National Aerospace Laboratory ATTN: SAVAA-D P.O. Box 90502 FT MONMOUTH, NJ 07703-5401 1006 BM Amsterdam USA Netherlands AVIONICS PANEL EXECUTIVE LTC James E. CLAY, US Army Telephone Telex Telefax (33) (1) 47-38-57-65 610176 (33) (1) 47-38-57-99 MAILING ADDRESSES: From Europe and Canada From United States AGARD AGARD ATTN: AVIONICS PANEL ATTN: AVIONICS PANEL 7, rue Ancelle APO NY 09777 92200 Neuilly-sur-Seine France ATTACHMENT 1 FOR US AUTHORS ONLY 1. Authors of US papers involving work performed or sponsored by a US Government Agency must receive clearance from their sponsoring agency. These authors should allow at least six weeks for clearance from their sponsoring agency. Abstracts, notices of clearance by sponsoring agencies, and Attachment 2 should be sent to Mr GAGGIN to arrive not later than 15 AUGUST 1990. 2. All other US authors should forward abstracts and Attachment 2 to Mr GAGGIN to arrive before 31 JULY 1990. These contributors should include the following statements in the cover letter: A. The work described was not performed under sponsorship of a US Government Agency. B. The abstract is technically correct. C. The abstract is unclassified. D. The abstract does not violate any proprietary rights. 3. US authors should send their abstracts to Mr GAGGIn and Dr KRUEGER only. Abstracts should NOT be sent to non-US members of the Technical Programme Committee or the Avionics Panel Executive. ABSTRACTS OF PAPERS FROM US AUTHORS CAN ONLY BE SENT TO: Mr David V. GAGGIN and Dr Charles H. KRUEGER Jr Director Director, Avionics Systems Div Avionics Research & Dev Activity Wright Research & Dev Center ATTN: SAVAA-D ATTN: WRDC/AAA Ft Monmouth, NJ 07703-5401 Wright Patterson AFB Dayton, OH 45433 Telephone: (201) 544-4851 Telephone: (513) 255-5218 or AUTOVON: 995-4851 4. US authors should send the Author Information Form (Attachment 2) to the Avionics Panel Executive, Mr GAGGIN, Dr KRUEGER, and each Technical Programme Committee Member, to meet the above deadlines. 5. Authors selected from the United States are remined that their full papers must be cleared by an authorized national clearance office before they can be forwarded to AGARD. Clearance procedures should be started at least 12 weeks before the paper is to be mailed to AGARD. Mr GAGGIN will provide additional information at the appropriate time. AUTHOR INFORMATION FORM FOR AUTHORS SUBMITTING AN ABSTRACT FOR THE AVIONICS PANEL SYMPOSIUM on MACHINE INTELLIGENCE FOR AEROSPACE ELECTRONICS SYSTEMS INSTRUCTIONS 1. Authors should complete this form and send a copy to the Avionics Panel Executive and all Technical Program Committee members by 31 AUGUST 1990. 2. Attach a copy of your abstract to these forms before they are mailed. US Authors must comply with ATTACHMENT 1 requirements. a. Probable Title Paper: __________________________________________ _____________________________________________________________________ b. Paper most appropriate for Session # ____________________________ c. Full Name of Author to be listed first on Programmee, including Courtesy Title, First Name and/or Initials, Last Name & Nationality. d. Name of Organization or Activity: _______________________________ _____________________________________________________________________ e. Address for Return Correspondence: Telephone Number: __________________________________ __________________ __________________________________ Telefax Number: __________________________________ __________________ __________________________________ Telex Number: __________________________________ __________________ f. Names of Co-Authors including Courtesy Titles, First Name and/or Initials, Last Name, their Organization, and their nationality. _________________________________________________________________ _________________________________________________________________ _________________________________________________________________ _________________________________________________________________ _________________________________________________________________ __________ ____________________ Date Signature DUE NOT LATER THAN 21 SEPTEMBER 1990 ------------------------------ End of Neuron Digest [Volume 6 Issue 49] ****************************************