Average System Performance Evaluation using Markov Chain

Size: px
Start display at page:

Download "Average System Performance Evaluation using Markov Chain"

Transcription

1 Designing high performant Systems: statistical timing analysis and optimization Average System Performance Evaluation using Markov Chain Weiyun Lu Supervisor: Martin Radetzki Sommer Semester 2006 Stuttgart University

2 Overview Scope & Motivation Example of a Labelled Transition System Markov Chain Performance Evaluation with Markov Chain Summary 2

3 Scope & Motivation System level Motivation -- Complexity of systems -- High impact of system-level decision System Modelling -- Property checking & Performance Evaluation Performance Modelling and Evaluation [1] Example of bus-based or switch based system for communiation 3

4 Overview Scope & Motivation Example of a Labelled Transition System Markov Chain Performance Evaluation with Markov Chain Summary 4

5 SHESim window of a Protocal Stack Lossy Channel Example -- SHE: Software/Hardware Engineering -- Model-driven, object-oriented framework for complex system specification. -- UML(class diagram, sequence diagram...) POOSL definition of a Lossy Channel -- POOSL: Parallel Object-Oriented Specification Language -- Formal: models are executable [2] transferframes()() f: Frame in?frame(f); if errordistribution yieldssuccess then delay(transmissiontime); out!frame(f); transferframe()() else transferframe()() fi. 5

6 transferframes()() f: Frame in?frame(f); Transition System of Lossy Channel if errordistribution yieldssuccess then delay(transmissiontime); out!frame(f); transferframe()() Bernoulli, Uniform, DiscreteUniform, Exponential, Normal distribution are available now in POOSL else fi. transferframe()() S1 out S4 Labeled Transition System (S, Λ, ): S: States; Λ: Labels; Transitions in S2 τ S3 3 6

7 Another Representation of Lossy Channel Step1: Assume maximal Progress Environment Modeling: -- 'open' system: (sub-modules) the environment is always willing to participate. -- 'closed' system: (complete system) no possibility to interact with environment Step2: Resolving non-determinism Difference between non-determinism & possibility transition Step3: Shifting Action and Time Information into States -- To simplify labels of transition: occurrence of actions and passage of time can be shifted into states. -- So the state becomes (S, e), e here can be an action or time passage S1, - in S1 S2 τ out S4 S3 S1, out S4, S1, τ S2, in S3, τ

8 ... (A) sel skip;... (B) or if (ErrorDistribution yieldssucess) then... (C) else... fi;... (D) or Method()() (E) les;... B A Resolving Non-Determinism Step2: Resolving non-determinism τ E D C A External Scheduler is needed! SHESim uses uniform distribution for POOSL Models Otherwise transform to Markov Decision Process 1/3 1/3 B 1/30 3/10 E D C 8

9 Overview Scope & Motivation Example of a Labelled Transition System Markov Chain Performance Evaluation with Markov Chain Summary 9

10 Markov Chain Discrete stochastic process: -- a sequence of probability events, {X i,i 0} -- 'i' here is called time-epoch, if it is in time domain. Markov Property: transitions only depend on current states Markov Chain & Representations Time-homogeneous (stationary) Markov Chain PA, A P A, B P A,C P A, D P A,E P B, A P B, B P B,C P B, D P B,E P row = P C, A P C, B P C,C P C, D P C, E = P D, A P D, B P D,C P D, D P D, E T P col = P row P E, A P E, B P E,C P E, D P E, E S1, - S1, out S4, 3 S1, τ 0.1 S2, in S3, τ A D E C B 10

11 Define initial distribution as Markov Chain We are interested in long-run average time fraction in each state of the Markov Chain. u 0 = u A0 u B0 u C0 u D0 u E0 Define at time epoch i, the probability to stay in each state as u i = u Ai u Bi u Ci u Di u Ei u A1 =u A0 P A, A u B0 P B, A u C0 P C, A u D0 P D, A u E0 P E, A <--- Conditional probability u 1 =u 0 P row u i 1 =u i P row <--- Generalize n u n =u 0 P row <--- Stationary 11

12 Markov Chain Compation for the Lossy Channel example: u 0 = u 1 = u 2 = u 0 = Calculate the long-run average time fraction in each state: u 50 = u 51 = u = 1 2 u 2n u 2n 1 = u 100 = u 101 = u 1000 = u 1001 = u 2n u 2n

13 Ergodic Markov Chain Ergodic Markov chain: -- Def1: possible to go from every state to every state, not necessarily in one move. -- Def2: some power of the the transition matrix has only positive entries. Equilibrium distribution: -- it can be proved in mathematics that for an ergodic Markov Chain with transition probability matrix P, there exists a unique probability vector u, such that u P = u -- u is called equilibrium distribution and is strictly positive. -- u denotes the long-run average time fraction of Markov chain in each state. For matrix of lossy channel u is calculated as u = ( 5/19 9/38 9/38 9/38 1/38 ) = ( ) n For Def2, it can be proved that: u 0 P n Ergodic Markov chain: Def3: if it has a positive state which is reachable from any other states with probability 1. In this case u is non-negative. u 13

14 Overview Scope & Motivation Example of a Labelled Transition System Markov Chain Performance Evaluation with Markov Chain Summary 14

15 Reward Function of Lossy Channel Reward function -- A funcion r : S --> R or {True(1), false(0)}, defined for a Markov Chain with state space S. -- Each time a state is visited, a reward specified by reward function is obtained. S1, - S1, out S4, 3 S1, τ 0.1 S2, in S3, τ 0.9 out(s,e) = { 1, if e = out 0, otherwise in(s,e) = { 1, if e = in 0, otherwise t(s,e) = { e, if e є (R) 0, otherwise out=0 in=0 t=0 out=1 in=0 t=0 out=0 in=1 t=0 0.1 out=0 in=0 t=0 0.9 out=0 in=0 t=3 out=0 in=0 t=0 15

16 Several Performance metrics of Lossy Channel Long-run average performance metric: For Lossy Channel, long-run average of : (i) number of in actions performed per time epoch of the Markov chain. (ii) number of in actions performed per unit of model time. (Capacity) (iii) time between two in actions. (iv) variance of time between two in actions. lim n n 1 n i=1 r X i Simple metrics: (i) (ii) Complex metrics: (iii) (iv) can be deduced from Atomic rewards only deducible from accumulated atomic rewards 16

17 Ergodic Theorem Ergodic Theorem n 1 n i=1 r X i a.s. T S u T r X T a.s. : almost surely: when n goes to infinity, with probability 1 r is a properly defined reward function for Markoc chain {X i,i 0} 17

18 Performance Evaluation of Lossy Channel long-run average of : (i) number of in actions performed per discrete-time epoch of the Markov chain. C1 = 1 n n i=1 (ii) number of in actions performed per unit of model time. C2 = S0:0 out=0 in=0 t=0 (iii) The long-run average time between two in actions. C3 = 1 n i=1 1 C2 n in S,e C1 t S,e = = 5 19 = 5/19 3 9/38 = D: 9/38 out=1 in=0 t=0 out=0 in=1 t=0 out=0 in=0 t=0 E:1/ C:9/38 out=0 in=0 t=3 out=0 in=0 t=0 A: 5/19 B:9/38 18

19 Classical Performance Evaluation Techniques (assume: an ergodic markov Chain with a proper reward function is defined.) Analytical method: Calculate equilibrium distribution u using u P=u, then apply ergodic theorem. Simulation-based method: lim n n 1 n i=1 r X i --Trace: a finite state sequence S = (S1, S2,....Sn ). n is the length of the trace. The larger is n, the more accurate is the result. ( see slide 11, regarded as r(s)=1. ) -- Point estimation θ: to calculate a single value θ on base of sample data.this value is to serve as a "best guess" for the unknown parameter θ. -- Interval estimation: to give a accuracy bound of a point estimation θ. [φ1, φ2] P(θ [ φ1, φ2 ] ) ß, ß [0, 1], ß is called confidence level 19

20 What have been presented Summary -- Transforming a labelled transition system into a Markoc Chain ( Lossy Channel, 3 steps ) -- Markoc Chain ( Ergodic Markov Chain & Equilibrium Distribution ) -- Performance Evaluation with Markoc Chain ( Reward Function & Ergodic Theorem, Analytical and Simulation-based method ) 20

21 Further Topics Temporal rewards in [2] Markov Chain reduction in [1] <--- for complex or delay-type metrics <--- conditional metrics SHE tools <--- relative strong tool for control software, formal framework for computation software and hardware synthesis are still under research ref[1]: Bart Theelen. Performance modelling for system-level design. Technische Universitaet Eindhoven, PhD Thesis, ref[2]: Jeroen P.M. Voeten. Performance evaluation with temporal rewards. Performance Evaluation, 50: ,

22 Andreyevich Markov ( ) End!

Development of dynamically evolving and self-adaptive software. 1. Background

Development of dynamically evolving and self-adaptive software. 1. Background Development of dynamically evolving and self-adaptive software 1. Background LASER 2013 Isola d Elba, September 2013 Carlo Ghezzi Politecnico di Milano Deep-SE Group @ DEIB 1 Requirements Functional requirements

More information

Master s Theory Exam Spring 2006

Master s Theory Exam Spring 2006 Spring 2006 This exam contains 7 questions. You should attempt them all. Each question is divided into parts to help lead you through the material. You should attempt to complete as much of each problem

More information

Random access protocols for channel access. Markov chains and their stability. Laurent Massoulié.

Random access protocols for channel access. Markov chains and their stability. Laurent Massoulié. Random access protocols for channel access Markov chains and their stability [email protected] Aloha: the first random access protocol for channel access [Abramson, Hawaii 70] Goal: allow machines

More information

Introduction to Markov Chain Monte Carlo

Introduction to Markov Chain Monte Carlo Introduction to Markov Chain Monte Carlo Monte Carlo: sample from a distribution to estimate the distribution to compute max, mean Markov Chain Monte Carlo: sampling using local information Generic problem

More information

TD(0) Leads to Better Policies than Approximate Value Iteration

TD(0) Leads to Better Policies than Approximate Value Iteration TD(0) Leads to Better Policies than Approximate Value Iteration Benjamin Van Roy Management Science and Engineering and Electrical Engineering Stanford University Stanford, CA 94305 [email protected] Abstract

More information

1. (First passage/hitting times/gambler s ruin problem:) Suppose that X has a discrete state space and let i be a fixed state. Let

1. (First passage/hitting times/gambler s ruin problem:) Suppose that X has a discrete state space and let i be a fixed state. Let Copyright c 2009 by Karl Sigman 1 Stopping Times 1.1 Stopping Times: Definition Given a stochastic process X = {X n : n 0}, a random time τ is a discrete random variable on the same probability space as

More information

Reinforcement Learning

Reinforcement Learning Reinforcement Learning LU 2 - Markov Decision Problems and Dynamic Programming Dr. Martin Lauer AG Maschinelles Lernen und Natürlichsprachliche Systeme Albert-Ludwigs-Universität Freiburg [email protected]

More information

IEOR 6711: Stochastic Models, I Fall 2012, Professor Whitt, Final Exam SOLUTIONS

IEOR 6711: Stochastic Models, I Fall 2012, Professor Whitt, Final Exam SOLUTIONS IEOR 6711: Stochastic Models, I Fall 2012, Professor Whitt, Final Exam SOLUTIONS There are four questions, each with several parts. 1. Customers Coming to an Automatic Teller Machine (ATM) (30 points)

More information

Software Performance and Scalability

Software Performance and Scalability Software Performance and Scalability A Quantitative Approach Henry H. Liu ^ IEEE )computer society WILEY A JOHN WILEY & SONS, INC., PUBLICATION Contents PREFACE ACKNOWLEDGMENTS xv xxi Introduction 1 Performance

More information

Managing Overloaded Hosts for Dynamic Consolidation of Virtual Machines in Cloud Data Centers Under Quality of Service Constraints

Managing Overloaded Hosts for Dynamic Consolidation of Virtual Machines in Cloud Data Centers Under Quality of Service Constraints IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, VOL. 24, NO. 7, JULY 2013 1366 Managing Overloaded Hosts for Dynamic Consolidation of Virtual Machines in Cloud Data Centers Under Quality of Service

More information

E3: PROBABILITY AND STATISTICS lecture notes

E3: PROBABILITY AND STATISTICS lecture notes E3: PROBABILITY AND STATISTICS lecture notes 2 Contents 1 PROBABILITY THEORY 7 1.1 Experiments and random events............................ 7 1.2 Certain event. Impossible event............................

More information

FEGYVERNEKI SÁNDOR, PROBABILITY THEORY AND MATHEmATICAL

FEGYVERNEKI SÁNDOR, PROBABILITY THEORY AND MATHEmATICAL FEGYVERNEKI SÁNDOR, PROBABILITY THEORY AND MATHEmATICAL STATIsTICs 4 IV. RANDOm VECTORs 1. JOINTLY DIsTRIBUTED RANDOm VARIABLEs If are two rom variables defined on the same sample space we define the joint

More information

Single item inventory control under periodic review and a minimum order quantity

Single item inventory control under periodic review and a minimum order quantity Single item inventory control under periodic review and a minimum order quantity G. P. Kiesmüller, A.G. de Kok, S. Dabia Faculty of Technology Management, Technische Universiteit Eindhoven, P.O. Box 513,

More information

Reinforcement Learning

Reinforcement Learning Reinforcement Learning LU 2 - Markov Decision Problems and Dynamic Programming Dr. Joschka Bödecker AG Maschinelles Lernen und Natürlichsprachliche Systeme Albert-Ludwigs-Universität Freiburg [email protected]

More information

Operations Research and Financial Engineering. Courses

Operations Research and Financial Engineering. Courses Operations Research and Financial Engineering Courses ORF 504/FIN 504 Financial Econometrics Professor Jianqing Fan This course covers econometric and statistical methods as applied to finance. Topics

More information

Business Process Modeling

Business Process Modeling Business Process Concepts Process Mining Kelly Rosa Braghetto Instituto de Matemática e Estatística Universidade de São Paulo [email protected] January 30, 2009 1 / 41 Business Process Concepts Process

More information

Theorem (informal statement): There are no extendible methods in David Chalmers s sense unless P = NP.

Theorem (informal statement): There are no extendible methods in David Chalmers s sense unless P = NP. Theorem (informal statement): There are no extendible methods in David Chalmers s sense unless P = NP. Explication: In his paper, The Singularity: A philosophical analysis, David Chalmers defines an extendible

More information

Exam Introduction Mathematical Finance and Insurance

Exam Introduction Mathematical Finance and Insurance Exam Introduction Mathematical Finance and Insurance Date: January 8, 2013. Duration: 3 hours. This is a closed-book exam. The exam does not use scrap cards. Simple calculators are allowed. The questions

More information

1/3 1/3 1/3 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0 1 2 3 4 5 6 7 8 0.6 0.6 0.6 0.6 0.6 0.6 0.6

1/3 1/3 1/3 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0 1 2 3 4 5 6 7 8 0.6 0.6 0.6 0.6 0.6 0.6 0.6 HOMEWORK 4: SOLUTIONS. 2. A Markov chain with state space {, 2, 3} has transition probability matrix /3 /3 /3 P = 0 /2 /2 0 0 Show that state 3 is absorbing and, starting from state, find the expected

More information

Stochastic Processes and Queueing Theory used in Cloud Computer Performance Simulations

Stochastic Processes and Queueing Theory used in Cloud Computer Performance Simulations 56 Stochastic Processes and Queueing Theory used in Cloud Computer Performance Simulations Stochastic Processes and Queueing Theory used in Cloud Computer Performance Simulations Florin-Cătălin ENACHE

More information

Overview of Violations of the Basic Assumptions in the Classical Normal Linear Regression Model

Overview of Violations of the Basic Assumptions in the Classical Normal Linear Regression Model Overview of Violations of the Basic Assumptions in the Classical Normal Linear Regression Model 1 September 004 A. Introduction and assumptions The classical normal linear regression model can be written

More information

Probability and statistics; Rehearsal for pattern recognition

Probability and statistics; Rehearsal for pattern recognition Probability and statistics; Rehearsal for pattern recognition Václav Hlaváč Czech Technical University in Prague Faculty of Electrical Engineering, Department of Cybernetics Center for Machine Perception

More information

Load Balancing and Switch Scheduling

Load Balancing and Switch Scheduling EE384Y Project Final Report Load Balancing and Switch Scheduling Xiangheng Liu Department of Electrical Engineering Stanford University, Stanford CA 94305 Email: [email protected] Abstract Load

More information

CITY UNIVERSITY LONDON. BEng Degree in Computer Systems Engineering Part II BSc Degree in Computer Systems Engineering Part III PART 2 EXAMINATION

CITY UNIVERSITY LONDON. BEng Degree in Computer Systems Engineering Part II BSc Degree in Computer Systems Engineering Part III PART 2 EXAMINATION No: CITY UNIVERSITY LONDON BEng Degree in Computer Systems Engineering Part II BSc Degree in Computer Systems Engineering Part III PART 2 EXAMINATION ENGINEERING MATHEMATICS 2 (resit) EX2005 Date: August

More information

PCHS ALGEBRA PLACEMENT TEST

PCHS ALGEBRA PLACEMENT TEST MATHEMATICS Students must pass all math courses with a C or better to advance to the next math level. Only classes passed with a C or better will count towards meeting college entrance requirements. If

More information

Nonparametric adaptive age replacement with a one-cycle criterion

Nonparametric adaptive age replacement with a one-cycle criterion Nonparametric adaptive age replacement with a one-cycle criterion P. Coolen-Schrijner, F.P.A. Coolen Department of Mathematical Sciences University of Durham, Durham, DH1 3LE, UK e-mail: [email protected]

More information

2014-2015 The Master s Degree with Thesis Course Descriptions in Industrial Engineering

2014-2015 The Master s Degree with Thesis Course Descriptions in Industrial Engineering 2014-2015 The Master s Degree with Thesis Course Descriptions in Industrial Engineering Compulsory Courses IENG540 Optimization Models and Algorithms In the course important deterministic optimization

More information

Lecture 7: Continuous Random Variables

Lecture 7: Continuous Random Variables Lecture 7: Continuous Random Variables 21 September 2005 1 Our First Continuous Random Variable The back of the lecture hall is roughly 10 meters across. Suppose it were exactly 10 meters, and consider

More information

Big Data Technology Motivating NoSQL Databases: Computing Page Importance Metrics at Crawl Time

Big Data Technology Motivating NoSQL Databases: Computing Page Importance Metrics at Crawl Time Big Data Technology Motivating NoSQL Databases: Computing Page Importance Metrics at Crawl Time Edward Bortnikov & Ronny Lempel Yahoo! Labs, Haifa Class Outline Link-based page importance measures Why

More information

Master of Arts in Mathematics

Master of Arts in Mathematics Master of Arts in Mathematics Administrative Unit The program is administered by the Office of Graduate Studies and Research through the Faculty of Mathematics and Mathematics Education, Department of

More information

The Relation between Two Present Value Formulae

The Relation between Two Present Value Formulae James Ciecka, Gary Skoog, and Gerald Martin. 009. The Relation between Two Present Value Formulae. Journal of Legal Economics 15(): pp. 61-74. The Relation between Two Present Value Formulae James E. Ciecka,

More information

LECTURE 4. Last time: Lecture outline

LECTURE 4. Last time: Lecture outline LECTURE 4 Last time: Types of convergence Weak Law of Large Numbers Strong Law of Large Numbers Asymptotic Equipartition Property Lecture outline Stochastic processes Markov chains Entropy rate Random

More information

3.2 Roulette and Markov Chains

3.2 Roulette and Markov Chains 238 CHAPTER 3. DISCRETE DYNAMICAL SYSTEMS WITH MANY VARIABLES 3.2 Roulette and Markov Chains In this section we will be discussing an application of systems of recursion equations called Markov Chains.

More information

1 Portfolio Selection

1 Portfolio Selection COS 5: Theoretical Machine Learning Lecturer: Rob Schapire Lecture # Scribe: Nadia Heninger April 8, 008 Portfolio Selection Last time we discussed our model of the stock market N stocks start on day with

More information

Chapter 1. Introduction

Chapter 1. Introduction Chapter 1 Introduction 1.1. Motivation Network performance analysis, and the underlying queueing theory, was born at the beginning of the 20th Century when two Scandinavian engineers, Erlang 1 and Engset

More information

CoolaData Predictive Analytics

CoolaData Predictive Analytics CoolaData Predictive Analytics 9 3 6 About CoolaData CoolaData empowers online companies to become proactive and predictive without having to develop, store, manage or monitor data themselves. It is an

More information

Alessandro Birolini. ineerin. Theory and Practice. Fifth edition. With 140 Figures, 60 Tables, 120 Examples, and 50 Problems.

Alessandro Birolini. ineerin. Theory and Practice. Fifth edition. With 140 Figures, 60 Tables, 120 Examples, and 50 Problems. Alessandro Birolini Re ia i it En ineerin Theory and Practice Fifth edition With 140 Figures, 60 Tables, 120 Examples, and 50 Problems ~ Springer Contents 1 Basic Concepts, Quality and Reliability Assurance

More information

How will the programme be delivered (e.g. inter-institutional, summerschools, lectures, placement, rotations, on-line etc.):

How will the programme be delivered (e.g. inter-institutional, summerschools, lectures, placement, rotations, on-line etc.): Titles of Programme: Hamilton Hamilton Institute Institute Structured PhD Structured PhD Minimum 30 credits. 15 of Programme which must be obtained from Generic/Transferable skills modules and 15 from

More information

Exercises in Mathematical Analysis I

Exercises in Mathematical Analysis I Università di Tor Vergata Dipartimento di Ingegneria Civile ed Ingegneria Informatica Eercises in Mathematical Analysis I Alberto Berretti, Fabio Ciolli Fundamentals Polynomial inequalities Solve the

More information

Discrete-Event Simulation

Discrete-Event Simulation Discrete-Event Simulation Prateek Sharma Abstract: Simulation can be regarded as the emulation of the behavior of a real-world system over an interval of time. The process of simulation relies upon the

More information

10-601. Machine Learning. http://www.cs.cmu.edu/afs/cs/academic/class/10601-f10/index.html

10-601. Machine Learning. http://www.cs.cmu.edu/afs/cs/academic/class/10601-f10/index.html 10-601 Machine Learning http://www.cs.cmu.edu/afs/cs/academic/class/10601-f10/index.html Course data All up-to-date info is on the course web page: http://www.cs.cmu.edu/afs/cs/academic/class/10601-f10/index.html

More information

MATH4427 Notebook 2 Spring 2016. 2 MATH4427 Notebook 2 3. 2.1 Definitions and Examples... 3. 2.2 Performance Measures for Estimators...

MATH4427 Notebook 2 Spring 2016. 2 MATH4427 Notebook 2 3. 2.1 Definitions and Examples... 3. 2.2 Performance Measures for Estimators... MATH4427 Notebook 2 Spring 2016 prepared by Professor Jenny Baglivo c Copyright 2009-2016 by Jenny A. Baglivo. All Rights Reserved. Contents 2 MATH4427 Notebook 2 3 2.1 Definitions and Examples...................................

More information

An Extension Model of Financially-balanced Bonus-Malus System

An Extension Model of Financially-balanced Bonus-Malus System An Extension Model of Financially-balanced Bonus-Malus System Other : Ratemaking, Experience Rating XIAO, Yugu Center for Applied Statistics, Renmin University of China, Beijing, 00872, P.R. China Phone:

More information

4.1. Title: data analysis (systems analysis). 4.2. Annotation of educational discipline: educational discipline includes in itself the mastery of the

4.1. Title: data analysis (systems analysis). 4.2. Annotation of educational discipline: educational discipline includes in itself the mastery of the 4.1. Title: data analysis (systems analysis). 4.4. Term of study: 7th semester. 4.1. Title: data analysis (applied mathematics). 4.4. Term of study: 6th semester. 4.1. Title: data analysis (computer science).

More information

Computational Learning Theory Spring Semester, 2003/4. Lecture 1: March 2

Computational Learning Theory Spring Semester, 2003/4. Lecture 1: March 2 Computational Learning Theory Spring Semester, 2003/4 Lecture 1: March 2 Lecturer: Yishay Mansour Scribe: Gur Yaari, Idan Szpektor 1.1 Introduction Several fields in computer science and economics are

More information

Continued Fractions and the Euclidean Algorithm

Continued Fractions and the Euclidean Algorithm Continued Fractions and the Euclidean Algorithm Lecture notes prepared for MATH 326, Spring 997 Department of Mathematics and Statistics University at Albany William F Hammond Table of Contents Introduction

More information

Hydrodynamic Limits of Randomized Load Balancing Networks

Hydrodynamic Limits of Randomized Load Balancing Networks Hydrodynamic Limits of Randomized Load Balancing Networks Kavita Ramanan and Mohammadreza Aghajani Brown University Stochastic Networks and Stochastic Geometry a conference in honour of François Baccelli

More information

General Theory of Differential Equations Sections 2.8, 3.1-3.2, 4.1

General Theory of Differential Equations Sections 2.8, 3.1-3.2, 4.1 A B I L E N E C H R I S T I A N U N I V E R S I T Y Department of Mathematics General Theory of Differential Equations Sections 2.8, 3.1-3.2, 4.1 Dr. John Ehrke Department of Mathematics Fall 2012 Questions

More information

Notes V General Equilibrium: Positive Theory. 1 Walrasian Equilibrium and Excess Demand

Notes V General Equilibrium: Positive Theory. 1 Walrasian Equilibrium and Excess Demand Notes V General Equilibrium: Positive Theory In this lecture we go on considering a general equilibrium model of a private ownership economy. In contrast to the Notes IV, we focus on positive issues such

More information

Vilnius University. Faculty of Mathematics and Informatics. Gintautas Bareikis

Vilnius University. Faculty of Mathematics and Informatics. Gintautas Bareikis Vilnius University Faculty of Mathematics and Informatics Gintautas Bareikis CONTENT Chapter 1. SIMPLE AND COMPOUND INTEREST 1.1 Simple interest......................................................................

More information

Gambling with Information Theory

Gambling with Information Theory Gambling with Information Theory Govert Verkes University of Amsterdam January 27, 2016 1 / 22 How do you bet? Private noisy channel transmitting results while you can still bet, correct transmission(p)

More information

Binomial lattice model for stock prices

Binomial lattice model for stock prices Copyright c 2007 by Karl Sigman Binomial lattice model for stock prices Here we model the price of a stock in discrete time by a Markov chain of the recursive form S n+ S n Y n+, n 0, where the {Y i }

More information

A Uniform Asymptotic Estimate for Discounted Aggregate Claims with Subexponential Tails

A Uniform Asymptotic Estimate for Discounted Aggregate Claims with Subexponential Tails 12th International Congress on Insurance: Mathematics and Economics July 16-18, 2008 A Uniform Asymptotic Estimate for Discounted Aggregate Claims with Subexponential Tails XUEMIAO HAO (Based on a joint

More information

The Analysis of Dynamical Queueing Systems (Background)

The Analysis of Dynamical Queueing Systems (Background) The Analysis of Dynamical Queueing Systems (Background) Technological innovations are creating new types of communication systems. During the 20 th century, we saw the evolution of electronic communication

More information

Some Polynomial Theorems. John Kennedy Mathematics Department Santa Monica College 1900 Pico Blvd. Santa Monica, CA 90405 [email protected].

Some Polynomial Theorems. John Kennedy Mathematics Department Santa Monica College 1900 Pico Blvd. Santa Monica, CA 90405 rkennedy@ix.netcom. Some Polynomial Theorems by John Kennedy Mathematics Department Santa Monica College 1900 Pico Blvd. Santa Monica, CA 90405 [email protected] This paper contains a collection of 31 theorems, lemmas,

More information

Chapter 6. The stacking ensemble approach

Chapter 6. The stacking ensemble approach 82 This chapter proposes the stacking ensemble approach for combining different data mining classifiers to get better performance. Other combination techniques like voting, bagging etc are also described

More information

ECE 842 Report Implementation of Elliptic Curve Cryptography

ECE 842 Report Implementation of Elliptic Curve Cryptography ECE 842 Report Implementation of Elliptic Curve Cryptography Wei-Yang Lin December 15, 2004 Abstract The aim of this report is to illustrate the issues in implementing a practical elliptic curve cryptographic

More information

ISU Department of Mathematics. Graduate Examination Policies and Procedures

ISU Department of Mathematics. Graduate Examination Policies and Procedures ISU Department of Mathematics Graduate Examination Policies and Procedures There are four primary criteria to be used in evaluating competence on written or oral exams. 1. Knowledge Has the student demonstrated

More information

Conductance, the Normalized Laplacian, and Cheeger s Inequality

Conductance, the Normalized Laplacian, and Cheeger s Inequality Spectral Graph Theory Lecture 6 Conductance, the Normalized Laplacian, and Cheeger s Inequality Daniel A. Spielman September 21, 2015 Disclaimer These notes are not necessarily an accurate representation

More information

Some Research Problems in Uncertainty Theory

Some Research Problems in Uncertainty Theory Journal of Uncertain Systems Vol.3, No.1, pp.3-10, 2009 Online at: www.jus.org.uk Some Research Problems in Uncertainty Theory aoding Liu Uncertainty Theory Laboratory, Department of Mathematical Sciences

More information

Introduction to Probability

Introduction to Probability Introduction to Probability EE 179, Lecture 15, Handout #24 Probability theory gives a mathematical characterization for experiments with random outcomes. coin toss life of lightbulb binary data sequence

More information

The Basics of Graphical Models

The Basics of Graphical Models The Basics of Graphical Models David M. Blei Columbia University October 3, 2015 Introduction These notes follow Chapter 2 of An Introduction to Probabilistic Graphical Models by Michael Jordan. Many figures

More information

Psychology and Economics (Lecture 17)

Psychology and Economics (Lecture 17) Psychology and Economics (Lecture 17) Xavier Gabaix April 13, 2004 Vast body of experimental evidence, demonstrates that discount rates are higher in the short-run than in the long-run. Consider a final

More information

A Profit-Maximizing Production Lot sizing Decision Model with Stochastic Demand

A Profit-Maximizing Production Lot sizing Decision Model with Stochastic Demand A Profit-Maximizing Production Lot sizing Decision Model with Stochastic Demand Kizito Paul Mubiru Department of Mechanical and Production Engineering Kyambogo University, Uganda Abstract - Demand uncertainty

More information

Black-box Performance Models for Virtualized Web. Danilo Ardagna, Mara Tanelli, Marco Lovera, Li Zhang [email protected]

Black-box Performance Models for Virtualized Web. Danilo Ardagna, Mara Tanelli, Marco Lovera, Li Zhang ardagna@elet.polimi.it Black-box Performance Models for Virtualized Web Service Applications Danilo Ardagna, Mara Tanelli, Marco Lovera, Li Zhang [email protected] Reference scenario 2 Virtualization, proposed in early

More information

Markovian Process and Novel Secure Algorithm for Big Data in Two-Hop Wireless Networks

Markovian Process and Novel Secure Algorithm for Big Data in Two-Hop Wireless Networks Markovian Process and Novel Secure Algorithm for Big Data in Two-Hop Wireless Networks K. Thiagarajan, Department of Mathematics, PSNA College of Engineering and Technology, Dindigul, India. A. Veeraiah,

More information

Statistics Graduate Courses

Statistics Graduate Courses Statistics Graduate Courses STAT 7002--Topics in Statistics-Biological/Physical/Mathematics (cr.arr.).organized study of selected topics. Subjects and earnable credit may vary from semester to semester.

More information

Dynamics and Equilibria

Dynamics and Equilibria Dynamics and Equilibria Sergiu Hart Presidential Address, GAMES 2008 (July 2008) Revised and Expanded (November 2009) Revised (2010, 2011, 2012, 2013) SERGIU HART c 2008 p. 1 DYNAMICS AND EQUILIBRIA Sergiu

More information

FULL LIST OF REFEREED JOURNAL PUBLICATIONS Qihe Tang

FULL LIST OF REFEREED JOURNAL PUBLICATIONS Qihe Tang FULL LIST OF REFEREED JOURNAL PUBLICATIONS Qihe Tang 87. Li, J.; Tang, Q. Interplay of insurance and financial risks in a discrete-time model with strongly regular variation. Bernoulli 21 (2015), no. 3,

More information

SPARE PARTS INVENTORY SYSTEMS UNDER AN INCREASING FAILURE RATE DEMAND INTERVAL DISTRIBUTION

SPARE PARTS INVENTORY SYSTEMS UNDER AN INCREASING FAILURE RATE DEMAND INTERVAL DISTRIBUTION SPARE PARS INVENORY SYSEMS UNDER AN INCREASING FAILURE RAE DEMAND INERVAL DISRIBUION Safa Saidane 1, M. Zied Babai 2, M. Salah Aguir 3, Ouajdi Korbaa 4 1 National School of Computer Sciences (unisia),

More information

EXERCISES FOR THE COURSE MATH 570, FALL 2010

EXERCISES FOR THE COURSE MATH 570, FALL 2010 EXERCISES FOR THE COURSE MATH 570, FALL 2010 EYAL Z. GOREN (1) Let G be a group and H Z(G) a subgroup such that G/H is cyclic. Prove that G is abelian. Conclude that every group of order p 2 (p a prime

More information

Homework set 4 - Solutions

Homework set 4 - Solutions Homework set 4 - Solutions Math 495 Renato Feres Problems R for continuous time Markov chains The sequence of random variables of a Markov chain may represent the states of a random system recorded at

More information

6.207/14.15: Networks Lecture 15: Repeated Games and Cooperation

6.207/14.15: Networks Lecture 15: Repeated Games and Cooperation 6.207/14.15: Networks Lecture 15: Repeated Games and Cooperation Daron Acemoglu and Asu Ozdaglar MIT November 2, 2009 1 Introduction Outline The problem of cooperation Finitely-repeated prisoner s dilemma

More information

Note on some explicit formulae for twin prime counting function

Note on some explicit formulae for twin prime counting function Notes on Number Theory and Discrete Mathematics Vol. 9, 03, No., 43 48 Note on some explicit formulae for twin prime counting function Mladen Vassilev-Missana 5 V. Hugo Str., 4 Sofia, Bulgaria e-mail:

More information

1 Short Introduction to Time Series

1 Short Introduction to Time Series ECONOMICS 7344, Spring 202 Bent E. Sørensen January 24, 202 Short Introduction to Time Series A time series is a collection of stochastic variables x,.., x t,.., x T indexed by an integer value t. The

More information

DATA ANALYSIS II. Matrix Algorithms

DATA ANALYSIS II. Matrix Algorithms DATA ANALYSIS II Matrix Algorithms Similarity Matrix Given a dataset D = {x i }, i=1,..,n consisting of n points in R d, let A denote the n n symmetric similarity matrix between the points, given as where

More information

Chapter 6: Sensitivity Analysis

Chapter 6: Sensitivity Analysis Chapter 6: Sensitivity Analysis Suppose that you have just completed a linear programming solution which will have a major impact on your company, such as determining how much to increase the overall production

More information

Module1. x 1000. y 800.

Module1. x 1000. y 800. Module1 1 Welcome to the first module of the course. It is indeed an exciting event to share with you the subject that has lot to offer both from theoretical side and practical aspects. To begin with,

More information

From the probabilities that the company uses to move drivers from state to state the next year, we get the following transition matrix:

From the probabilities that the company uses to move drivers from state to state the next year, we get the following transition matrix: MAT 121 Solutions to Take-Home Exam 2 Problem 1 Car Insurance a) The 3 states in this Markov Chain correspond to the 3 groups used by the insurance company to classify their drivers: G 0, G 1, and G 2

More information

Random graphs with a given degree sequence

Random graphs with a given degree sequence Sourav Chatterjee (NYU) Persi Diaconis (Stanford) Allan Sly (Microsoft) Let G be an undirected simple graph on n vertices. Let d 1,..., d n be the degrees of the vertices of G arranged in descending order.

More information

Background Knowledge

Background Knowledge Background Knowledge Precalculus GEOMETRY Successful completion of the course with a grade of B or higher Solid understanding of: Right Triangles Congruence Theorems Basic Trigonometry Basic understanding

More information

Analysis of a Production/Inventory System with Multiple Retailers

Analysis of a Production/Inventory System with Multiple Retailers Analysis of a Production/Inventory System with Multiple Retailers Ann M. Noblesse 1, Robert N. Boute 1,2, Marc R. Lambrecht 1, Benny Van Houdt 3 1 Research Center for Operations Management, University

More information

How To Balance In A Distributed System

How To Balance In A Distributed System 6 IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, VOL. 11, NO. 1, JANUARY 2000 How Useful Is Old Information? Michael Mitzenmacher AbstractÐWe consider the problem of load balancing in dynamic distributed

More information

Notes from Week 1: Algorithms for sequential prediction

Notes from Week 1: Algorithms for sequential prediction CS 683 Learning, Games, and Electronic Markets Spring 2007 Notes from Week 1: Algorithms for sequential prediction Instructor: Robert Kleinberg 22-26 Jan 2007 1 Introduction In this course we will be looking

More information

Computing Near Optimal Strategies for Stochastic Investment Planning Problems

Computing Near Optimal Strategies for Stochastic Investment Planning Problems Computing Near Optimal Strategies for Stochastic Investment Planning Problems Milos Hauskrecfat 1, Gopal Pandurangan 1,2 and Eli Upfal 1,2 Computer Science Department, Box 1910 Brown University Providence,

More information