An Algorithm for Bootstrapping Communications Jun Wang 03/20/03.

Slides:



Advertisements
Similar presentations
Intelligent Technologies Module: Ontologies and their use in Information Systems Revision lecture Alex Poulovassilis November/December 2009.
Advertisements

Automata Theory Part 1: Introduction & NFA November 2002.
Pseudo-Relevance Feedback For Multimedia Retrieval By Rong Yan, Alexander G. and Rong Jin Mwangi S. Kariuki
1 Minimally Supervised Morphological Analysis by Multimodal Alignment David Yarowsky and Richard Wicentowski.
1 CD5560 FABER Formal Languages, Automata and Models of Computation Lecture 2 Mälardalen University 2005.
1 1 CDT314 FABER Formal Languages, Automata and Models of Computation Lecture 3 School of Innovation, Design and Engineering Mälardalen University 2012.
1.1 Aims of testing Testing is a process centered around the goal of finding defects in a system. We are currently unable to produce defect-free systems.
Image Analysis Phases Image pre-processing –Noise suppression, linear and non-linear filters, deconvolution, etc. Image segmentation –Detection of objects.
1 Module 20 NFA’s with -transitions –NFA- ’s Formal definition Simplifies construction –LNFA- –Showing LNFA  is a subset of LNFA (extra credit) and therefore.
B.Macukow 1 Lecture 12 Neural Networks. B.Macukow 2 Neural Networks for Matrix Algebra Problems.
Kostas Kontogiannis E&CE
Module 14 Thought & Language. INTRODUCTION Definitions –Cognitive approach method of studying how we process, store, and use information and how this.
Content Based Image Clustering and Image Retrieval Using Multiple Instance Learning Using Multiple Instance Learning Xin Chen Advisor: Chengcui Zhang Department.
Chapter 4 DECISION SUPPORT AND ARTIFICIAL INTELLIGENCE
1 Introduction to Computability Theory Lecture12: Reductions Prof. Amos Israeli.
1 Module 2: Fundamental Concepts Problems Programs –Programming languages.
Neural Networks Basic concepts ArchitectureOperation.
Lecture 2: Fundamental Concepts
1 Module 2: Fundamental Concepts Problems Programs –Programming languages.
Applications of Evolutionary Computation in the Analysis of Factors Influencing the Evolution of Human Language Alex Decker.
Un Supervised Learning & Self Organizing Maps Learning From Examples
Lecture 18 NFA’s with -transitions –NFA- ’s Formal definition Simplifies construction –LNFA- –Showing LNFA  is a subset of LNFA and therefore a subset.
Gene Regulatory Networks - the Boolean Approach Andrey Zhdanov Based on the papers by Tatsuya Akutsu et al and others.
MACHINE LEARNING. What is learning? A computer program learns if it improves its performance at some task through experience (T. Mitchell, 1997) A computer.
Overview  Description of ELLs  Obstacles in Math for ELLs  Help ELLs in Math  My Critique.
Chapter 10 Systems Planning, Analysis, and Design.
MSE 2400 EaLiCaRA Spring 2015 Dr. Tom Way
Web Services Experience Language Web Services eXperience Language Technical Overview Ravi Konuru e-Business Tools and Frameworks,
Reinforcement Learning
CSE 311 Foundations of Computing I Lecture 21 Finite State Machines Autumn 2011 CSE 3111.
CSE 311 Foundations of Computing I Lecture 21 Finite State Machines Spring
By Ian Jackman Davit Stepanyan.  User executed untested code.  The order in which statements were meant to be executed are different than the order.
An Ontological Framework for Web Service Processes By Claus Pahl and Ronan Barrett.
Lesson Planning: part # 1 Lecture # 7. Review of Lesson # 6 We talked about the following elements of Presentation, Practice and Production stages of.
Equipping ourselves for the 21 st century Andreina España.
Effective Communication
1 CD5560 FABER Formal Languages, Automata and Models of Computation Lecture 3 Mälardalen University 2010.
Project quality management. Introduction Project quality management includes the process required to ensure that the project satisfies the needs for which.
Learning to Share Meaning in a Multi-Agent System (Part I) Ganesh Padmanabhan.
Progression of Learning Elementary English as a Second Language MELS Working Document 2009.
1 Lecture 6 Neural Network Training. 2 Neural Network Training Network training is basic to establishing the functional relationship between the inputs.
CSE 589 Part V One of the symptoms of an approaching nervous breakdown is the belief that one’s work is terribly important. Bertrand Russell.
1 Statistics & R, TiP, 2011/12 Neural Networks  Technique for discrimination & regression problems  More mathematical theoretical foundation  Works.
1. 2 »Requires following laws and proper procedures »Requires people with strong human relation and communication skills »Responsibilities include: –maintaining.
The Challenge of Amorphous Computing To develop engineering principles and programming techniques for directing the behavior of systems composed of billions.
Program Implementation. Consulting Roles Analyst Change Agent –Systems –People Trainer Yes Man Listener/Counselor Scapegoat.
Onlinedeeneislam.blogspot.com1 Design and Analysis of Algorithms Slide # 1 Download From
Organizational Culture & Environment
Listening Reading Speaking Interacting It concerns focuses on language features such as pronunciation, spelling, collocations, etc. Major Effects:
Artificial Intelligence Knowledge Representation.
CLIL: Methodology and Applications Team work: Mazzarelli Gioconda, Plenzick Angelina, Vaccarella Lucia, Vertucci Italia. Liceo Scientifico G. Rummo – BN.
Unit 1: Present Tense   Simple Present Tense   Present Continuous Tense   Subject & Object Pronouns (I, you, it, he, she, they) vs. (me, you, him,
Directed Evolution of a Genetic Circuit 15 February 2008 George McArthur This presentation is primarily based on content taken from the paper “Directed.
Victoria Ibarra Mat:  Generally, Computer hardware is divided into four main functional areas. These are:  Input devices Input devices  Output.
       January 3 rd, 2005 The signaling properties of the individual neuron. How we move from understanding individual nerve cells.
Information Systems Development
Organizational Behavior (MGT-502)
Probability Vocabulary
Modeling Arithmetic, Computation, and Languages
Pushdown Automata PDAs
Universal Systems Model
Information Systems Development
Motivation Computers are good at some things… Calculating 
Competitive Networks.
A synthetic multicellular system for programmed pattern formation
EA C461 – Artificial Intelligence Problem Solving Agents
Competitive Networks.
CSE 311: Foundations of Computing
Presentation transcript:

An Algorithm for Bootstrapping Communications Jun Wang 03/20/03

Amorphous Computing Amorphous computing is the development of organizational principles and programming languages for obtaining coherent behavior from the cooperation of myriads of unreliable parts that are interconnected in un known, irregular, and time-varying ways.

Aim of amorphous computing Structuring systems so we get acceptable answers, with high probability, even in the face of unreliability

Motivation “If I were designing the human brain, how would I have the parts learn to communicate?” A kind of amorphous computing

A possible solution A,B: two agents Communication Lines: a bundle of wires with an arbitrary and unknown permutation Feature Lines: connection with outside world

Work mechanism Communication lines have four states: 1, -1, 0, x; Feature lines are named by things or actions and driven by roles; Typical feature lines: bob, mary, push … Typical roles: subject, object, verb…

Performance Evaluation Method Training cycles – data on the feature lines is sent to both agents; Test cycles – one agent has no input. How well the output of it matches the values of the other agent’s feature line?

Encoding View

An Example (1)

An Example (2)

An Example (3) A symbol is expressed by a subset of communication wires; symbol mapping: (x s,x c,x u,x n ) Inflection mapping: (role, value)

Algorithm Analysis Talk-in Talk outListen in Listen out Basic idea: each agent transmits the mapping information to other by driving the communication lines, and modify his own mapping information according to the other’s mapping information. After some cycles, the two agents will get the same symbol mapping.

Formal Automation Description(1) Talk in: Adding new elements into two mapping relation; updating communication lines states; Talk out: Driving the communication lines

Formal Automation Description(2) Listen in: (1) Symbol mapping adjustment: two agents listen to each other in every cycle and modify the “certain” subsets. Finally, they can reach to a same symbol mapping eventually; (2) Preparing the output for listen-out; (3) Inflection mapping adjustment; Listen out: output the (symbol,role) pairs to feature lines according to the understanding.

Results N w = 10000; N wps = 100; Training cycles = 1000; Thematic frames: 50 nouns, 20 verbs Results: 200 cycles for share vocabulary; 500 cycles for inflection stability

Algorithm feature Structural parsimony Robustness Shallow computation

Performance Degradation

Dissimilar Features The algorithm can share vocabulary despite the handicaps imposed like some non-shared vocabulary between them.