Applications of Multi-Agent Learning in E-Commerce and Autonomic Computing Jeff Kephart IBM Research December 14, 2002.

Slides:



Advertisements
Similar presentations
Tivoli Software from IBM Storage Resource Management Webcast
Advertisements

KAIS T The Vision of Autonomic Computing Jeffrey O. Kephart, David M Chess IBM Watson research Center IEEE Computer, Jan 발표자 : 이승학.
Welcome to DEAS 2005 Design and Evolution of Autonomic Application Software David Garlan, CMU Marin Litoiu, IBM CAS Hausi A. Müller, UVic John Mylopoulos,
Presented by: Ehsan Kouroshfar. Outline Introduction Self-management Architectural Considerations Engineering challenges Scientific challenges Discussion.
CROWN “Thales” project Optimal ContRol of self-Organized Wireless Networks WP1 Understanding and influencing uncoordinated interactions of autonomic wireless.
Autonomic Systems Justin Moles, Winter 2006 Security in an Autonomic Computing Environment Paper by: D. M. Chess, C. C. Palmer S. R. White Presentation.
The Role of Prices in Peer-Assisted Content Distribution Christina Aperijis Michael J. Freedman Ramesh Johari Presented by: Kyle Chauvin and Henry Xie.
Chapter 1: Expanding Abroad Motivations, Means, and Mentalities
Fabián E. Bustamante, Winter 2006 Autonomic Computing The vision of autonomic computing, J. Kephart and D. Chess, IEEE Computer, Jan Also - A.G.
Sogang University ICC Lab Using Game Theory to Analyze Wireless Ad Hoc networks.
SysMA S OFTWARE E NGINEERING FOR S ERVICE - O RIENTED AND A UTONOMIC S YSTEMS - I NTRODUCTION TO A UTONOMIC S YSTEMS - Lucca (Italy) – October, 2014 Francesco.
Yingping Huang and Gregory Madey University of Notre Dame A W S utonomic eb-based imulation Presented by Tariq M. King Published by the IEEE Computer Society.
Effective Coordination of Multiple Intelligent Agents for Command and Control The Robotics Institute Carnegie Mellon University PI: Katia Sycara
January 18, 2005 Experimental Economics 1 Outline  Introduction: Objectives and Course Expectation  Double Oral Auction Experiment: Observations  Foundations.
A Heuristic Bidding Strategy for Multiple Heterogeneous Auctions Patricia Anthony & Nicholas R. Jennings Dept. of Electronics and Computer Science University.
ISA5428: 普及計算 Autonomic Computing 金仲達教授 清華大學資訊系統與應用研究所 九十一學年度第二學期 (Slides are taken from the presentations by Alan Ganek, Alfred Spector, Jeff Kephart.
1 A General Auction-Based Architecture for Resource Allocation Weidong Cui, Matthew C. Caesar, and Randy H. Katz EECS, UC Berkeley {wdc, mccaesar,
Adaptive Infrastructures EPRI/DoD Initiative on Complex Interactive Networks/Systems Joint innovative research ·EPRI and ·Office of the Director of Defense.
InfoCenters and Information E-markets Itai Yarom PhD Researcher - AI Lab
Managing the Information Technology Resource Jerry N. Luftman
Improving Robustness in Distributed Systems Jeremy Russell Software Engineering Honours Project.
Autonomic Computing Shafay Shamail Malik Jahan Khan.
SING* and ToNC * Scientific Foundations for Internet’s Next Generation Sirin Tekinay Program Director Theoretical Foundations Communication Research National.
ChemConnect Leading in negotiation solutions for commercial products.
Introduction to the new mainframe: Large-Scale Commercial Computing © Copyright IBM Corp., All rights reserved. Chapter 8: Autonomic computing.
(Slides are taken from the presentations by Alan Ganek, Alfred Spector, Jeff Kephart of IBM)
26th May, Middleware or Simulator for Autonomic Communications Yang Qiu Networking Laboratory Helsinki University of Technology
Relationship Marketing
Demonstrating IT Relevance to Business Aligning IT and Business Goals with On Demand Automation Solutions Robert LeBlanc General Manager Tivoli Software.
Computer System Lifecycle Chapter 1. Introduction Computer System users, administrators, and designers are all interested in performance evaluation. Whether.
Accelerating Product and Service Innovation © 2013 IBM Corporation IBM Integrated Solution for System z Development (ISDz) Henk van der Wijk 23 Januari.
A Concise Introduction to Autonomic Computing
IT – DBMS Concepts Relational Database Theory.
Trading Agent Competition (Supply Chain Management) and TacTex-05.
1 Autonomic Computing An Introduction Guenter Kickinger.
Distributed Real-Time Systems for the Intelligent Power Grid Prof. Vincenzo Liberatore.
Supply Contract Allocation Gyana R. Parija Bala Ramachandran IBM T.J. Watson Research Center INFORMS Miami 2001.
Professor Ran Kivetz Columbia Business School Spring 2007 High Technology Marketing and Entrepreneurship THE ROLE OF MARKETING (& “Can this High-tech Product.
WELCOME. AUTONOMIC COMPUTING PRESENTED BY: NIKHIL P S7 IT ROLL NO: 33.
Chapter 6 Sourcing. Objectives After reading the chapter and reviewing the materials presented the students will be able to: Explain the difference between.
A Lightweight Platform for Integration of Resource Limited Devices into Pervasive Grids Stavros Isaiadis and Vladimir Getov University of Westminster
Living markets ® living agents ® Adaptive Execution in Business Networks January 21 st, 2002.
What are the main differences and commonalities between the IS and DA systems? How information is transferred between tasks: (i) IS it may be often achieved.
1. Process Gather Input – Today Form Coherent Consensus – Next two months.
CompSci Self-Managing Systems Shivnath Babu.
IBM Global Services © Copyright IBM Corporation 2005 International Business Machines ITIL Capacity Management Deep Dive Chris Molloy IBM Distinguished.
1 Intrusion Detection Methods “Intrusion detection is the process of identifying and responding to malicious activity targeted at computing and networking.
1 ACTIVE FAULT TOLERANT SYSTEM for OPEN DISTRIBUTED COMPUTING (Autonomic and Trusted Computing 2006) Giray Kömürcü.
9 Systems Analysis and Design in a Changing World, Fourth Edition.
THE VISION OF AUTONOMIC COMPUTING. WHAT IS AUTONOMIC COMPUTING ? “ Autonomic Computing refers to computing infrastructure that adapts (automatically)
Basic Principles of Economics Rögnvaldur J. Sæmundsson January
Paul A. Strassmann, Copyright Stevens Institute of Technology The Structure of I.T. Spending as Measure of Organizational Disorder Paul A. Strassmann,
Algorithmic, Game-theoretic and Logical Foundations
Managing Web Server Performance with AutoTune Agents by Y. Diao, J. L. Hellerstein, S. Parekh, J. P. Bigus Presented by Changha Lee.
The Vision of Autonomic Computing Self-Management Unit 7-2 Managing the Digital Enterprise Kephart, and Chess.
Chapter 8 System Management Semester 2. Objectives  Evaluating an operating system  Cooperation among components  The role of memory, processor,
Lecture 2Hayek Hypothesis and Institution as a Variable "Markets as Economizers of Information: Experimental Examination of the ‘Hayek Hypothesis’," Economic.
Application Communities Phase 2 (AC2) Project Overview Nov. 20, 2008 Greg Sullivan BAE Systems Advanced Information Technologies (AIT)
A Study on Policy-Based Interaction Techniques with Autonomic Computing Peter Khooshabeh University of California, Santa Barbara 1 Department of Psychology,
Multi-Agents System CMSC 691B Gunjan Kalra Peter DSouza.
AUTONOMIC COMPUTING B.Akhila Priya 06211A0504. Present-day IT environments are complex, heterogeneous in terms of software and hardware from multiple.
Computational Economy
Database Management System (DBMS)
For modeling conflict and cooperation Schwartz/Teneketzis
Research Challenges of Autonomic Computing
The Vision of Autonomic Computing
Celian Colon (Ecole Polytechnique, Ecole Normale supérieure, Paris)
Topic:- Autonomic Computing Presented By :- Neera Gupta
Presentation transcript:

Applications of Multi-Agent Learning in E-Commerce and Autonomic Computing Jeff Kephart IBM Research December 14, 2002

Two broad application areas E-commerce Large-scale competitive MAS Billions of economically motivated agents Buying and selling information goods and services Adaptive, and coupled directly and indirectly (through markets) Autonomic computing Large-scale cooperative or competitive MAS Self-managing computing systems Self-configuring, Self-healing, Self-optimizing, Self-protecting IEEE Computer, January 2003

The Future Information Economy Billions of interacting, adaptive agents. What emergent behaviors will arise?

Experiment Bidding Agents vs. Humans CDA is common in financial markets Extensive prior literature –All-human experiments (Vernon Smith) –All-agent experiments (SFI DA, Gode-Sunder, Cliff, TAC) Market History Time Price Trade Ask Bid Continuous Double Auction

Watson Experimental Economics Lab Copyright New York Times

Seller GUI (CDA) Limit Prices Bid Queue Ask Queue Submit

Wakeup? Compute Order Brain Bookkeeper Message Handler Market state Agent state Bidding Agent Place Order Bidding Agent Architecture Auction Info Orders Auctioneer Agent GUI

Agent-Human experiments Human subjects recruited from local colleges and IBM Research given interactive instructions and test paid in proportion to surplus Setup 6 Humans, 6 Agents 6 Buyers, 6 Sellers Each agent shares limit prices with a human Experiment 9 to 16 3-minute periods Limit prices change every 3-5 periods Record bids, asks, trades

Experiment #6: Fast GD vs. Humans

Summary of experimental results Agents won by substantial margins in all experiments ~20% more surplus than novice humans ~5-7% more surplus than experienced humans Agents and humans interact with one another Not two decoupled markets ~30-50% of trades are agent-human Market efficiency improves with number of agents Humans fare better when there are more agents Agents can supplant humans as economic decision makers

All-agent experiments Simulator Discrete-time; stochastic asynchronous dynamics. Ran mixtures of several strategies and variants Market parameters 10 buyers, 10 sellers 10 units each. Fixed limit prices (chosen randomly) 100 expts, 5 trading pds/expt, 300 time steps/pd. Experimental comparisons Homogeneous (0 A vs. 20 B) One-in-Many Tests (1 A vs. 19 B) Balanced Team Tests (10 A vs. 10 B)

Expt. 2: 1 A vs. 19 B Differential Efficiency ZIP and MGD invade ZI, Kaplan & GD But don’t invade one another Kaplan can invade all strategies All strategies invade ZI ZI doesn’t invade any

Expt. 3: 10 A vs. 10 B Differential Surplus (out of 2612 total) ZI beats Kaplan 100-0! Other strategies beat Kaplan, but by smaller margin GDX > MGD > ZIP > GD > ZI > Kaplan

Evolutionary dynamics CDA game What happens when agents gradually switch to more successful strategies? No strategy is dominant. This is a useful view for Mechanism design Agent design

Dynamic pricing game Shopbots and pricebots

Dynamic Pricing Meta-payoff matrix

Evolutionary dynamics Dynamic pricing game NE Computation 5 Agents20 Agents

Dynamic pricing game Shopbots and pricebots

A cure for myopia

Price response curves Symmetric and Asymmetric solutions Higher efficiency, but unstable!

Sequential Learning Limit Cycle

Autonomic Computing Self-managing computing systems Administration of individual systems is increasingly difficult 100s of configuration, tuning parameters for DB2, WebSphere Heterogeneous systems are becoming increasingly connected Integration becoming ever more difficult Architects can't intricately plan interactions among components Increasingly dynamic; more frequently with unanticipated components More of the burden must be assumed at run time But human system administrators can't assume the burden; already 6:1 cost ratio between storage admin and storage 40% outages due to operator error We need self-managing computing systems Behavior specified by sys admins via high-level policies System and its components figure out how to carry out policies

Evolving towards Self-management TodayThe Autonomic Future Self- configure Corporate data centers are multi-vendor, multi-platform. Installing, configuring, integrating systems is time- consuming, error-prone. Automated configuration of components, systems according to high-level policies; rest of system adjusts automatically. Seamless, like adding new cell to body or new individual to population. Self-heal Problem determination in large, complex systems can take a team of programmers weeks Automated detection, diagnosis, and repair of localized software/hardware problems. Self-optimize WebSphere, DB2 have hundreds of nonlinear tuning parameters; many new ones with each release. Components and systems will continually seek opportunities to improve their own performance and efficiency. Self-protect Manual detection and recovery from attacks and cascading failures. Automated defense against malicious attacks or cascading failures; use early warning to anticipate and prevent system-wide failures.

Autonomic element structure Fundamental atom of the architecture Managed element(s) Database, storage system, etc. Plus one autonomic manager Responsible for: Providing its service Managing its own behavior in accordance with policies Interacting with other autonomic elements An Autonomic Element Monitor Analyze Sensors Execute Plan Effectors Knowledge Autonomic Manager Managed Element SensorsEffectors

Autonomic elements interaction Relationships Dynamic, ephemeral Formed by agreement May be negotiated Full spectrum Peer-to-peer Hierarchical Subject to policies

AC System and Infrastructure SentinelBrokerAggregatorNegotiatorArbiterPlannerReputation Authority Registry Monitor SentinelNegotiatorArbiterBrokerEvent Correlator ServerDatabaseStorageNetworkServer ProvisionerWorkload Manager

Multi-agent Learning scenarios Designing system behavior Interacting feedback loops Negotiation and resource allocation Problem determination

Control & harness emergent behavior Understand, control, and exploit emergent behavior in autonomic systems. How do self-*, stability, etc. depend on Behaviors and goals of the agents Pattern and type of interactions among agents External influences and demands on system How to invert this relationship to achieve system goals?

Interacting control/optimization loops Transaction Requests Increase demand Server 1 DB Service Server 2 File System Storage Service 2 Storage Service 1 Increase service Feedback control & optimization of single autonomic elements Done for 1-2 variables What happens when feedback loops interact?

Interacting control/optimization loops Transaction Requests Increase demand Server 1 DB Service Server 2 File System Storage Service 2 Storage Service 1 Capacity limit reached: Get more storage X

Interacting control/optimization loops Transaction Requests Demand not being met: Find alternate supplier Server 1 DB Service Server 2 File System Storage Service 2 Storage Service 1 Getting more storage X

Interacting control/optimization loops Transaction Requests Sorry; already found an alternative Server 1 DB Service Server 2 File System Storage Service 2 Storage Service 1 Ready to give you that extra service X

Transaction Requests Server 1 DB 1 Server 2 File System 1 Storage 2Storage 1 Negotiation and resource allocation Request( QueryService, Queries = 800/sec, Type = 2, RT = 5 sec) Request( QueryService, Queries = 400/sec, Type = 5, RT = 3 sec) Request( TableSpace, Size = 3 GBytes, Reads = 2000/sec, Writes = 100/sec) Request( LogicalVolume, Size = 12 Gbytes, Reads = 500/sec, Writes = 500/sec) Counterpropose( TableSpace, Size = 3 GBytes, Reads = 1600/sec, Writes = 100/sec) Counterpropose( QueryService, Queries = 320/sec, Type = 5, RT = 4 sec) Policies: utility functions Compute costs, benefits from business contract, propagate them down. Forms of negotiation: Bilateral Multilateral Auction Supply chain Competitive/coop Learning During a negotiation Strategy evolution Collective behavior?

Problem Determination Construct adaptive statistical models of large networked systems Learn about inter-element dependencies (within locale) Determine model structure, parameters e.g. Bayes Net Monitor logs, use model to Detect potential problems Set up monitors as needed Diagnose problems Challenge Shifting topology

Closing remarks E-commerce: competitive, giga-agent MAS “Naïve” learning: apply single-agent learning, see what happens Analyze outcome Compute Nash eq. of meta-game, study evolutionary dynamics To improve strategies or improve market mechanisms Autonomic computing: cooperative or competitive MAS Coordinated learning is more possible Explicit correlation signals to reach correlated equilibria? Multi-agent Q learning Can dictate goals and strategies to agents IEEE Computer, January 2003

Backup slides

Autonomic Computing Architecture Based on a distributed, service-oriented architectural approach E.g., OGSA Every component provides or consumes services Policy-based management Autonomic elements Every component must be resilient, robust, self-managing Autonomic elements are the architecture’s way of achieving this for a component Behavior is specified and driven by policies Relationships between autonomic elements Based on agreements established and maintained by autonomic elements Governed by policies Give rise to resiliency, robustness, self-management of system Relationships are the architecture’s way of achieving this for system as a whole No architectural single point of failure

Engineering Challenge Conflict Resolution Increase throughput Decrease throughput Workload Manager Intrusion Detector Give me 700 MIPs 500 MBytes Give me 500 MIPs 600 MBytes Application Manager 1 Application Manager 2 Network Manager Server Element 1000 MIPs 1000 MBytes Priority(ID) > Priority(WLM) Utility(AM1) > Utility(AM2)

Engineering Challenge Human-Computer Interface Develop new languages, metaphors and translation technologies that enable humans to monitor, visualize, and control AC systems Specify goals and objectives to AC systems, and visualize their potential effect Techniques must be Sufficiently expressive of preferences regarding cost vs. performance, security, risk and reliability Sufficiently structured and/or naturally suited to human psychology and cognition to keep specification errors to an absolute minimum Robust to specification errors

Scientific Challenge Multi-agent Learning Today: Lots of good practical techniques for single agent to learn about a static agent or environment, with solid theory to back it up.

Scientific Challenge Multi-agent Learning Challenge: Establish theoretical foundation for understanding and performing learning and optimization in multi-agent systems.