Math 312, Intro. to Real Analysis: Homework #4 Solutions

Similar documents
Sequences and Series

The Limit of a Sequence (Brief Summary) 1

Solutions to Problem Sheet 1

5. Best Unbiased Estimators

Limits of sequences. Contents 1. Introduction 2 2. Some notation for sequences The behaviour of infinite sequences 3

Rafa l Kulik and Marc Raimondo. University of Ottawa and University of Sydney. Supplementary material

2.6 Rational Functions and Their Graphs

Further Pure 1 Revision Topic 5: Sums of Series

MATH 205 HOMEWORK #1 OFFICIAL SOLUTION

INTERVAL GAMES. and player 2 selects 1, then player 2 would give player 1 a payoff of, 1) = 0.

Maximum Empirical Likelihood Estimation (MELE)

Summary. Recap. Last Lecture. .1 If you know MLE of θ, can you also know MLE of τ(θ) for any function τ?

Notes on Expected Revenue from Auctions

14.30 Introduction to Statistical Methods in Economics Spring 2009

SELECTING THE NUMBER OF CHANGE-POINTS IN SEGMENTED LINE REGRESSION

A random variable is a variable whose value is a numerical outcome of a random phenomenon.

Combining imperfect data, and an introduction to data assimilation Ross Bannister, NCEO, September 2010

Fourier Transform in L p (R) Spaces, p 1

NORMALIZATION OF BEURLING GENERALIZED PRIMES WITH RIEMANN HYPOTHESIS

FINM6900 Finance Theory How Is Asymmetric Information Reflected in Asset Prices?

Overlapping Generations

Models of Asset Pricing

x satisfying all regularity conditions. Then

r i = a i + b i f b i = Cov[r i, f] The only parameters to be estimated for this model are a i 's, b i 's, σe 2 i

Statistics for Economics & Business

Section Mathematical Induction and Section Strong Induction and Well-Ordering

Lecture 9: The law of large numbers and central limit theorem

Chapter 8. Confidence Interval Estimation. Copyright 2015, 2012, 2009 Pearson Education, Inc. Chapter 8, Slide 1

Models of Asset Pricing

Models of Asset Pricing

EXERCISE - BINOMIAL THEOREM

Appendix 1 to Chapter 5

ECON 5350 Class Notes Maximum Likelihood Estimation

4.5 Generalized likelihood ratio test

A New Constructive Proof of Graham's Theorem and More New Classes of Functionally Complete Functions

STAT 135 Solutions to Homework 3: 30 points

of Asset Pricing R e = expected return

1 The Power of Compounding

1 Estimating sensitivities

Monetary Economics: Problem Set #5 Solutions

18.S096 Problem Set 5 Fall 2013 Volatility Modeling Due Date: 10/29/2013

Parametric Density Estimation: Maximum Likelihood Estimation

1 Random Variables and Key Statistics

SUPPLEMENTAL MATERIAL

EVEN NUMBERED EXERCISES IN CHAPTER 4

of Asset Pricing APPENDIX 1 TO CHAPTER EXPECTED RETURN APPLICATION Expected Return

Average Distance and Vertex-Connectivity

The material in this chapter is motivated by Experiment 9.

Today: Finish Chapter 9 (Sections 9.6 to 9.8 and 9.9 Lesson 3)

APPLICATION OF GEOMETRIC SEQUENCES AND SERIES: COMPOUND INTEREST AND ANNUITIES

Standard Deviations for Normal Sampling Distributions are: For proportions For means _

0.1 Valuation Formula:

MA Lesson 11 Section 1.3. Solving Applied Problems with Linear Equations of one Variable

point estimator a random variable (like P or X) whose values are used to estimate a population parameter

. The firm makes different types of furniture. Let x ( x1,..., x n. If the firm produces nothing it rents out the entire space and so has a profit of

Calculation of the Annual Equivalent Rate (AER)

We learned: $100 cash today is preferred over $100 a year from now

Mixed and Implicit Schemes Implicit Schemes. Exercise: Verify that ρ is unimodular: ρ = 1.

Game Theory. Lecture Notes By Y. Narahari. Department of Computer Science and Automation Indian Institute of Science Bangalore, India July 2012

When you click on Unit V in your course, you will see a TO DO LIST to assist you in starting your course.

Hopscotch and Explicit difference method for solving Black-Scholes PDE

Problem Set 1a - Oligopoly

Confidence Intervals. CI for a population mean (σ is known and n > 30 or the variable is normally distributed in the.

11.7 (TAYLOR SERIES) NAME: SOLUTIONS 31 July 2018

A Note About Maximum Likelihood Estimator in Hypergeometric Distribution

Unbiased estimators Estimators

Lecture 4: Parameter Estimation and Confidence Intervals. GENOME 560 Doug Fowler, GS

CHAPTER 8 Estimating with Confidence

1 + r. k=1. (1 + r) k = A r 1

The Top-Down Approach to Calculation of the Insurance Premium

Binomial Model. Stock Price Dynamics. The Key Idea Riskless Hedge

Chpt 5. Discrete Probability Distributions. 5-3 Mean, Variance, Standard Deviation, and Expectation

between 1 and 100. The teacher expected this task to take Guass several minutes to an hour to keep him busy but

Chapter 5: Sequences and Series

Online appendices from Counterparty Risk and Credit Value Adjustment a continuing challenge for global financial markets by Jon Gregory

CAPITAL PROJECT SCREENING AND SELECTION

CHAPTER : ARITHMETIC PROGRESSION CONTENTS. Idetify characteristics of arithmetic progressio PAGE 2.2 Determie whether a give sequece is a arithmetic p

1 Grim Trigger in the Repeated Prisoner s Dilemma (70 points)

MS-E2114 Investment Science Exercise 2/2016, Solutions

5 Statistical Inference

1 Basic Growth Models

Estimating Proportions with Confidence

Exam 2. Instructor: Cynthia Rudin TA: Dimitrios Bisias. October 25, 2011

Anomaly Correction by Optimal Trading Frequency

Marking Estimation of Petri Nets based on Partial Observation

Non-Inferiority Logrank Tests

On Regret and Options - A Game Theoretic Approach for Option Pricing

CHAPTER 8: CONFIDENCE INTERVAL ESTIMATES for Means and Proportions

ii. Interval estimation:

Outline. Plotting discrete-time signals. Sampling Process. Discrete-Time Signal Representations Important D-T Signals Digital Signals

LESSON #66 - SEQUENCES COMMON CORE ALGEBRA II

Budgetary Effects on Pricing Equilibrium in Online Markets

STRAND: FINANCE. Unit 3 Loans and Mortgages TEXT. Contents. Section. 3.1 Annual Percentage Rate (APR) 3.2 APR for Repayment of Loans

1 The Black-Scholes model

A DOUBLE INCREMENTAL AGGREGATED GRADIENT METHOD WITH LINEAR CONVERGENCE RATE FOR LARGE-SCALE OPTIMIZATION

arxiv: v3 [math.st] 3 May 2016

Forecasting bad debt losses using clustering algorithms and Markov chains

We consider the planning of production over the infinite horizon in a system with timevarying

Lecture 5 Point Es/mator and Sampling Distribu/on

CAPITAL ASSET PRICING MODEL

Transcription:

Math 3, Itro. to Real Aalysis: Homework #4 Solutios Stephe G. Simpso Moday, March, 009 The assigmet cosists of Exercises 0.6, 0.8, 0.0,.,.3,.6,.0,.,. i the Ross textbook. Each problem couts 0 poits. 0.6. (a) Let (s ) be a sequece of real umbers. The, wheever m is less tha we have s m s = (s m s m+ ) + (s m+ s m+ ) + + (s s ), hece by the Triagle Iequality s m s s m s m+ + s m+ s m+ + + s s. Now, if our sequece is such that s k s k+ / k for all k, the our iequality becomes s m s m + m+ + + ad the right-had side is the begiig of the geometrical series k=m /k = / m. Thus we see that s m s / m for all m ad all > m. Thus (s ) is a Cauchy sequece. Therefore, by oe of our theorems, (s ) is a coverget sequece. (b) If we assume oly that s k s k+ < /k for all k, we caot coclude that (s ) is a Cauchy sequece. A example illustratig this is obtaied by lettig s = the th partial sum of the harmoic series, s = H = + + 3 + +. For this example we have but s s + = + < lims = + + 3 + + + = so our sequece (s ) is diverget. = = + (harmoic series)

0.8. Assume that (s ) is a odecreasig sequece of real umbers. Let σ be the average of the first umbers i our give sequece: σ = s + + s. We claim that the sequece (σ ) is agai odecreasig. To see this, ote that s s s +, hece s + + s s +. Addig (s + + s ) to both sides, we obtai (s + + s ) + (s + + s ) (s + + s ) + s +, or, i other words, ( + )(s + + s ) (s + + s + s + ). Dividig both sides by ( + ), we obtai s + + s s + + s + s + + or, i other words, σ σ + for all. This proves our claim. The coclusio ca be restated as follows: If a sequece is odecreasig, the its sequece of averages is odecreasig. Note: The textbook assumes i additio that s is positive for all. However, this assumptio is ot eeded i order to obtai the coclusio. 0.0. Defie a sequece (s ) iductively by lettig s = ad s + = (s +)/3 for all. (a) The first few terms of the sequece are s = ad s = /3 ad s 3 = 5/9 ad s 4 = 4/7. (b) Obviously s > /. If s > /, it follows that s + > 3/, hece s + = (s + )/3 > (3/)/3, hece s > /. Thus by iductio o we see that s > / for all. (c) Sice s > /, it follows that s s + = s s + 3 = 3 s 3 > 3 3 = 0 so s > s + for all, i.e., (s ) is a decreasig sequece. (d) Sice (s ) is a decreasig sequece which is bouded below (the boud is /), it follows by oe of our theorems that (s ) is coverget. Let α = lims. Takig the limit of both sides of the equatio s + = (s + )/3, we obtai α = (α + )/3. Solvig for α we obtai α = /. Thus lims = /.

.. Let a = ( ), b =, c =, d = 6 + 4 7 3. Note that the sequece (a ) looks like,,,,.... (a) A mootoe subsequece of (a ) is,,,.... We ca describe this subsequece as (a k ) k where k = k. Thus a k = ( ) k = for all k. Alteratively, usig istead of k as the idex, we ca describe our subsequece as (a ). The sequeces (b ), (c ), ad (d ) are already mootoe. (Note that ay sequece is cosidered to be a subsequece of itself.) To see that (d ) is mootoe, use algebra to show directly that for all. 6 + 4 6( + ) + 4 > 7 3 7( + ) 3 (b) The subsequetial limits of our sequeces are ad for (a ), 0 for (b ), + for (c ), ad 6/7 for (d ). (c) limsup a =, limif a =. limsupb = limif b = limb = 0. limsup c = limif c = limc = +. limsupd = limif d = limd = 6/7. (d) (a ) is diverget ad its limit is udefied. (b ) ad (d ) are coverget to 0 ad 6/7 respectively. (c ) is diverget to +. (e) (a ), (b ), ad (d ) are bouded. (c ) is ubouded..3. Let s = cos π 3, t 3 = 4 +, ( u = ), v = ( ) +. Note that the sequece (s ) looks like,,,,,,,,,.... (a) A mootoe sbsequece of (s ) is,,,..., which ca be described as (s 6+3 ), i.e., (cos( + )π). The sequece (t ) is already mootoe. A mootoe subsequece of (u ) is (u ), i.e., (/ ), which looks like /4, /6, /64,.... A mootoe subsequece of (v ) is (v ), i.e., ( + /), which looks like 3/, 5/4, 7/6,.... (b) The subsequetial limits are ±, ±/ for (s ), 0 for (t ), 0 for (u ), ± for (v ). (c) (s ) ad (v ) are diverget ad their limits are udefied. (t ) ad (u ) are coverget to 0. (d) (s ), (t ), (u ), ad (v ) are bouded. 3

.6. Assume that (c ) is a subsequece of (b ) ad (b ) is a subsequece of (a ). By equatio (3) o page 64, there are fuctios f() ad g() such that b = a f() ad c = b g() for all. It follows that c = a f(g()) for all. Thus (c ) is a subsequece of (a ) via the composite fuctio f g..0. We are give the sequece, /,,, /, /3, /4, /3, /,,, /,.... The subsequetial limits of this sequece are, /, /3,..., /,... ad 0. The limsup is ad the limif is 0... By defiitio of limits, lims = 0 meas that for all ǫ > 0 we ca fid N so large that s 0 < ǫ for all > N. I other words, for all ǫ > 0 we ca fid N so large that s < ǫ for all > N. But the, by the defiitio of lim sup, this coditio is precisely equivalet to sayig that limsup s 0. Moreover, s 0 automatically, hece limsup s 0 automatically, for all sequeces. Thus, our coditio above is equivalet to sayig that limsup s = 0. We have ow proved that lims = 0 if ad oly if limsup s = 0... Assume that (s ) is a sequece of real umbers. As i Exercise 0.8, let σ be the average of s,..., s. (a) We shall prove the chai of iequalities limif s limif σ limsup σ limsup s. () The middle iequality limif σ limsup σ is obvious ad does ot require proof. We prove the right-had iequality limsup σ limsup s. We omit the proof of the left-had iequality limif s limif σ, which is similar. I order to prove limsupσ limsup s, we follow the hit give i the textbook. First, give M ad N such that M > N, we claim that sup σ s + + s N >M M To prove (), it suffices to prove σ s + + s N M + sup s k. () + sup s k (3) for each idividual > M. Sice > M > N, we ca break σ ito two parts: σ = s + + s = s + + s N + s N+ + + s. (4) Sice > M, the first part of the right-had side of (4) is < the first part of the right-had side of (3). Ad obviously, the secod part 4

of the right-had side of (4) is the secod part of the right-had side of (3). Combiig these two observatios, we see that (3) holds. From this it follows immediately that () holds. Now, fixig N ad takig the limit of () as M goes to ifiity, we see that limsup σ sup s k (5) for all N. Fially, takig the limit of (5) as N goes to ifiity, we see that limsup σ limsup s k ad this completes the proof. (b) I the case whe lims exists, we have lim s = limif s = limsups, so () becomes limif s = limif σ = limsupσ = limsups = lims. From this it follows that lim σ exists ad is equal to lims. This coclusio ca be restated as follows: If a sequece is coverget, the its sequece of averages is coverget. 5