Stanford Statistics 311/Electrical Engineering 377

Size: px
Start display at page:

Download "Stanford Statistics 311/Electrical Engineering 377"

Transcription

1 I. Uiversal predictio ad codig a. Gae: sequecex ofdata, adwattopredict(orcode)aswellasifwekew distributio of data b. Two versios: probabilistic ad adversarial. I either case, let p ad q be desities or probability ass fuctios (codes). Regret of sequece x is Reg(Q,P,x ) := log q(x ) log p(x ) = i= log q(x i x i ) log p(x i x i ). Associated axiu regret (usuall just regret) with respect to faily P is R X (Q,P) := sup Reg(Q,P,x ). P P,x X (Note i ore geerality, ca have loss fuctios up there other tha log-loss; will discuss ore later.) 2. Redudacy related to codig is expected regret uder distributio P. That is, Red (Q,P) := E P [log q(x ) log p(x ) = D kl (P Q). The worst-case redudacy with respect to a class P is R(Q,P) := sup Red (Q,P). P P We saw last tie that if q(x) = 2 l C(x), the we had the codig gae. c. Exaple: filterig proble. Suppose we believe X i N(AX i +g,σ 2 I), where we assue σ 2 is fixed ad kow. The we ight look at distributios Q that at iteratio i predict X i N(µ i,σ 2 I), i which case the two values are Reg(Q,P,x ) = i= 2σ 2 µ i x i 2 2 2σ 2 Ax i +g x i 2 2 for the regret, ad for the redudacy we have Red (Q,P) = 2σ 2 d. Miiax strategies for regret: i= E[ AX i +g µ i (X i ) Coplexity i the regret settig: assue we have paraetric set P = {P } Θ defied o X The coplexity of the set P is Cop (P) := log p (x )dx or geerally Cop (P) := log p (x )dµ(x ). X sup Θ 2. Actual value of iiax gae if Cop(P) < : X sup Θ

2 Propositio 0.. Assue that P has fiite coplexity Cop(P) <. The the iiax regret if Q RX (Q,P) = Cop (P) Proof Note that if we choose the oralized axiu likelihood distributio or Shtarkov distributio Q to have desity q(x ) = sup p (x )/ sup p (x )dx, the Q has costat regret [ R X (Q,P) = sup log x X = sup x [ log q(x ) log sup p (x ) sup p sup p (x ) log sup p (x ) = Cop (P). Moreover, for ay other distributio Q Q there is soe z X satisfyig q(z ) < q(z ), so that R X (Q,P) log q(z ) log sup p (z ) > log q(z ) log sup p (z ) = Cop (P), because Q assigs the sae probability to each sequece. 3. Exaple: Coplexity of the Beroulli distributio. We ay copute this early exactly. First, we paraeterize via [0,, so that for a sequece x {0,} with ozeros, we have for = / that ( ) ( ) P (x ) = ( ) = exp( h 2 ( )), where h 2 (p) = plogp ( p)log( p) is the biary etropy. Moreover, we have P (x ) = sup [0,P (x ). Thus we fid that ( ) Cop ([0,) = log e h 2( ). Now I cheat by usig Stirlig s approxiatio without tellig you: for ay p (0,) with p N, we have ( ) [, exp(h 2 (p)). p 8p( p) πp( p) =0 Dealig with = ad = 0 explicitly, we the obtai =0 ( )exp( h 2 ( )) = = [ 8, ( ) exp( h 2 ( )) π = ( }{{ ) } 0 (( )) 2 2

3 I particular, we have that as, (2+[8 /2,π /2 /2 if Q RX (Q,P) = Cop ([0,) = log (Fisher iforatio) II. Fisher iforatio = 2 log+log ( ) d +O(). ) d +o() ( ) a. Fisher Iforatio: let deote a paraetric faily, where is suitably sooth. I := E [ log log = E [ l l, where l = log. Note that with variables b. Alterate defiitios: Uder suitable soothess coditios, we have E [ l = logdx = dx = dx = dx = = 0. Also, because we have 2 log = 2 I = E [ l l = 2 = 2 l l, 2 logdx+ 2 dx = E[ 2 log+ 2 dx = E[ 2 log. } {{ } = c. A few exaple Fisher iforatios:. Exaple 0. (Caoical expoetial faily): We have log =,φ(x) A(), ad because l = φ(x) A() ad 2 log = 2 A(), we obtai I = 2 A(). 2. Exaple 0.2 (Two paraeterizatios of Beroulli): I caoical paraeterizatio froexpoetialfaily,wesawexp(x log(+e e )),sofisheriforatiois = +e +e p( p) uder chage of variables p = e /(+e ), or = log p p. O the other had, if P(X = x) = p x ( p) x, the logp(x = x) = x p x p, so that [ (X E p p X ) 2 = p p + p = p( p). 3

4 III. Fisher iforatio: Craér Rao Boud Propositio 0.2 (Craér Rao Boud). Let φ : R d R be arbitrary differetiable fuctio ad assue that T is ubiased for φ() uder P. The Var(T) φ() I φ(). Iediate corollary: take φ() = λ,, ad vary λ, ad we obtai that for ay ubiased estiator T for, Var( λ,t) λ I λ, or Cov(T) I. Proof First, if we ca ove derivatives i ad out wishy-washy-like, Cov(T φ(), l,j ) = E[(T φ()) l,j = E[T l,j = T(x) j dx = j T(x)dx = φ(). j Now, we ote that Var(T ) 0 for all λ, ad usig the previous equalities, Var(T ) = Var(T)+λ I 2E[T = Var(T)+λ I λ 2 λ, φ(). ) = Var(T) φ() I φ(), ad rear- Takig λ = I φ() gives 0 Var(T ragig gives the result. IV. Coectios of Fisher iforatio to divergeces a. Exaple 0.3(Divergeces i expoetial failies): Suppose = h(x) exp(, φ(x) A()). The D kl (P P 2 ) = A( 2 ) A( ) A( ), 2. Note that for sooth A, which is all expoetial faily odels, we thus have D kl (P P 2 ) = 2 2, 2 A( )( 2 ) +O( 2 3 ) = 2 2,I ( 2 )+O( 2 3 ). Coect with Brega divergece via B f (x,y) = f(x) f(y) f(y),x y. b. Other sketchy result: KL divergece ad Fisher iforatio. We clai that Propositio 0.3. For appropriately sooth failies, D kl (P P 2 ) = 2 2,I ( 2 )+o( 2 2 ). 4

5 Sketch of Proof We have E [logp 2 (X) = E [logp (X)+E [ logp (X), E [( 2 ) 2 logp (X)( 2 )+E [R(, 2,X), where R(, 2,X) is equal to the third derivative (tesor) evaluated at take at soe poit (X) = λ X + ( λ X ) 2, where λ X [0, ay deped o X, that is, R(, 2,X) = 6 3 (X)[ logp (X) 2 = O X ( 2 3 ), where O X deotes a big-oh ter depedig o X. Assuig that we ca ove derivatives outside of itegrals appropriately, we obtai E [logp 2 (X) = E [logp (X)+ p (x)dx, ( 2 ) E [ 2 logp (X)( 2 )+o( 2 2 ) = E [logp (X) 2 ( 2 ) I ( 2 )+o( 2 2 ). Rearragig ad otig that E [logp 2 (X) E [logp (X) = D kl (P P 2 ) gives the result. Reark: Coditios to ake all this go through are those sufficiet to apply Lebesgue s doiated covergece theore. So if, for the base easure µ we have the existece of a fuctio g such that g(x) f(x,) for all, where gdµ <, the f(x,)dµ = f(x,)dµ by the ea-value theore (ote that for all 0 we have the upper boud sup v: v 2 δ f(x,) =0 2 g(x)). More geerally, ca hadle absolutely cotiuous fuctios, which are differetiable alost +v everywhere. 5

A PROBABILITY PROBLEM

A PROBABILITY PROBLEM A PROBABILITY PROBLEM A big superarket chai has the followig policy: For every Euros you sped per buy, you ear oe poit (suppose, e.g., that = 3; i this case, if you sped 8.45 Euros, you get two poits,

More information

ECE 901 Lecture 4: Estimation of Lipschitz smooth functions

ECE 901 Lecture 4: Estimation of Lipschitz smooth functions ECE 9 Lecture 4: Estiatio of Lipschitz sooth fuctios R. Nowak 5/7/29 Cosider the followig settig. Let Y f (X) + W, where X is a rado variable (r.v.) o X [, ], W is a r.v. o Y R, idepedet of X ad satisfyig

More information

Probability Theory. Exercise Sheet 4. ETH Zurich HS 2017

Probability Theory. Exercise Sheet 4. ETH Zurich HS 2017 ETH Zurich HS 2017 D-MATH, D-PHYS Prof. A.-S. Szita Coordiator Yili Wag Probability Theory Exercise Sheet 4 Exercise 4.1 Let X ) N be a sequece of i.i.d. rado variables i a probability space Ω, A, P ).

More information

On Modeling On Minimum Description Length Modeling. M-closed

On Modeling On Minimum Description Length Modeling. M-closed O Modelig O Miiu Descriptio Legth Modelig M M-closed M-ope Do you believe that the data geeratig echais really is i your odel class M? 7 73 Miiu Descriptio Legth Priciple o-m-closed predictive iferece

More information

f(1), and so, if f is continuous, f(x) = f(1)x.

f(1), and so, if f is continuous, f(x) = f(1)x. 2.2.35: Let f be a additive fuctio. i Clearly fx = fx ad therefore f x = fx for all Z+ ad x R. Hece, for ay, Z +, f = f, ad so, if f is cotiuous, fx = fx. ii Suppose that f is bouded o soe o-epty ope set.

More information

Define a Markov chain on {1,..., 6} with transition probability matrix P =

Define a Markov chain on {1,..., 6} with transition probability matrix P = Pla Group Work 0. The title says it all Next Tie: MCMC ad Geeral-state Markov Chais Midter Exa: Tuesday 8 March i class Hoework 4 due Thursday Uless otherwise oted, let X be a irreducible, aperiodic Markov

More information

Statistical Theory MT 2009 Problems 1: Solution sketches

Statistical Theory MT 2009 Problems 1: Solution sketches Statistical Theory MT 009 Problems : Solutio sketches. Which of the followig desities are withi a expoetial family? Explai your reasoig. (a) Let 0 < θ < ad put f(x, θ) = ( θ)θ x ; x = 0,,,... (b) (c) where

More information

5.1 A mutual information bound based on metric entropy

5.1 A mutual information bound based on metric entropy Chapter 5 Global Fao Method I this chapter, we exted the techiques of Chapter 2.4 o Fao s method the local Fao method) to a more global costructio. I particular, we show that, rather tha costructig a local

More information

Unbiased Estimation. February 7-12, 2008

Unbiased Estimation. February 7-12, 2008 Ubiased Estimatio February 7-2, 2008 We begi with a sample X = (X,..., X ) of radom variables chose accordig to oe of a family of probabilities P θ where θ is elemet from the parameter space Θ. For radom

More information

EECS564 Estimation, Filtering, and Detection Hwk 2 Solns. Winter p θ (z) = (2θz + 1 θ), 0 z 1

EECS564 Estimation, Filtering, and Detection Hwk 2 Solns. Winter p θ (z) = (2θz + 1 θ), 0 z 1 EECS564 Estimatio, Filterig, ad Detectio Hwk 2 Sols. Witer 25 4. Let Z be a sigle observatio havig desity fuctio where. p (z) = (2z + ), z (a) Assumig that is a oradom parameter, fid ad plot the maximum

More information

Statistical Theory MT 2008 Problems 1: Solution sketches

Statistical Theory MT 2008 Problems 1: Solution sketches Statistical Theory MT 008 Problems : Solutio sketches. Which of the followig desities are withi a expoetial family? Explai your reasoig. a) Let 0 < θ < ad put fx, θ) = θ)θ x ; x = 0,,,... b) c) where α

More information

Fourier Series and their Applications

Fourier Series and their Applications Fourier Series ad their Applicatios The fuctios, cos x, si x, cos x, si x, are orthogoal over (, ). m cos mx cos xdx = m = m = = cos mx si xdx = for all m, { m si mx si xdx = m = I fact the fuctios satisfy

More information

Lecture 10: Universal coding and prediction

Lecture 10: Universal coding and prediction 0-704: Iformatio Processig ad Learig Sprig 0 Lecture 0: Uiversal codig ad predictio Lecturer: Aarti Sigh Scribes: Georg M. Goerg Disclaimer: These otes have ot bee subjected to the usual scrutiy reserved

More information

Chapter 6 Infinite Series

Chapter 6 Infinite Series Chapter 6 Ifiite Series I the previous chapter we cosidered itegrals which were improper i the sese that the iterval of itegratio was ubouded. I this chapter we are goig to discuss a topic which is somewhat

More information

IIT JAM Mathematical Statistics (MS) 2006 SECTION A

IIT JAM Mathematical Statistics (MS) 2006 SECTION A IIT JAM Mathematical Statistics (MS) 6 SECTION A. If a > for ad lim a / L >, the which of the followig series is ot coverget? (a) (b) (c) (d) (d) = = a = a = a a + / a lim a a / + = lim a / a / + = lim

More information

10/ Statistical Machine Learning Homework #1 Solutions

10/ Statistical Machine Learning Homework #1 Solutions Caregie Mello Uiversity Departet of Statistics & Data Sciece 0/36-70 Statistical Macie Learig Hoework # Solutios Proble [40 pts.] DUE: February, 08 Let X,..., X P were X i [0, ] ad P as desity p. Let p

More information

Chapter 6 Principles of Data Reduction

Chapter 6 Principles of Data Reduction Chapter 6 for BST 695: Special Topics i Statistical Theory. Kui Zhag, 0 Chapter 6 Priciples of Data Reductio Sectio 6. Itroductio Goal: To summarize or reduce the data X, X,, X to get iformatio about a

More information

Lecture 10: Bounded Linear Operators and Orthogonality in Hilbert Spaces

Lecture 10: Bounded Linear Operators and Orthogonality in Hilbert Spaces Lecture : Bouded Liear Operators ad Orthogoality i Hilbert Spaces 34 Bouded Liear Operator Let ( X, ), ( Y, ) i i be ored liear vector spaces ad { } X Y The, T is said to be bouded if a real uber c such

More information

Some Examples on Gibbs Sampling and Metropolis-Hastings methods

Some Examples on Gibbs Sampling and Metropolis-Hastings methods Soe Exaples o Gibbs Saplig ad Metropolis-Hastigs ethods S420/620 Itroductio to Statistical Theory, Fall 2012 Gibbs Sapler Saple a ultidiesioal probability distributio fro coditioal desities. Suppose d

More information

Solutions: Homework 3

Solutions: Homework 3 Solutios: Homework 3 Suppose that the radom variables Y,...,Y satisfy Y i = x i + " i : i =,..., IID where x,...,x R are fixed values ad ",...," Normal(0, )with R + kow. Fid ˆ = MLE( ). IND Solutio: Observe

More information

Statistics for Applications Fall Problem Set 7

Statistics for Applications Fall Problem Set 7 18.650. Statistics for Applicatios Fall 016. Proble Set 7 Due Friday, Oct. 8 at 1 oo Proble 1 QQ-plots Recall that the Laplace distributio with paraeter λ > 0 is the cotiuous probaλ bility easure with

More information

2. The volume of the solid of revolution generated by revolving the area bounded by the

2. The volume of the solid of revolution generated by revolving the area bounded by the IIT JAM Mathematical Statistics (MS) Solved Paper. A eigevector of the matrix M= ( ) is (a) ( ) (b) ( ) (c) ( ) (d) ( ) Solutio: (a) Eigevalue of M = ( ) is. x So, let x = ( y) be the eigevector. z (M

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 3 9/11/2013. Large deviations Theory. Cramér s Theorem

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 3 9/11/2013. Large deviations Theory. Cramér s Theorem MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/5.070J Fall 203 Lecture 3 9//203 Large deviatios Theory. Cramér s Theorem Cotet.. Cramér s Theorem. 2. Rate fuctio ad properties. 3. Chage of measure techique.

More information

Sequences and Series of Functions

Sequences and Series of Functions Chapter 6 Sequeces ad Series of Fuctios 6.1. Covergece of a Sequece of Fuctios Poitwise Covergece. Defiitio 6.1. Let, for each N, fuctio f : A R be defied. If, for each x A, the sequece (f (x)) coverges

More information

Convergence of random variables. (telegram style notes) P.J.C. Spreij

Convergence of random variables. (telegram style notes) P.J.C. Spreij Covergece of radom variables (telegram style otes).j.c. Spreij this versio: September 6, 2005 Itroductio As we kow, radom variables are by defiitio measurable fuctios o some uderlyig measurable space

More information

Lecture 8: Convergence of transformations and law of large numbers

Lecture 8: Convergence of transformations and law of large numbers Lecture 8: Covergece of trasformatios ad law of large umbers Trasformatio ad covergece Trasformatio is a importat tool i statistics. If X coverges to X i some sese, we ofte eed to check whether g(x ) coverges

More information

REVIEW OF CALCULUS Herman J. Bierens Pennsylvania State University (January 28, 2004) x 2., or x 1. x j. ' ' n i'1 x i well.,y 2

REVIEW OF CALCULUS Herman J. Bierens Pennsylvania State University (January 28, 2004) x 2., or x 1. x j. ' ' n i'1 x i well.,y 2 REVIEW OF CALCULUS Hera J. Bieres Pesylvaia State Uiversity (Jauary 28, 2004) 1. Suatio Let x 1,x 2,...,x e a sequece of uers. The su of these uers is usually deoted y x 1 % x 2 %...% x ' j x j, or x 1

More information

Statistics for Applications. Chapter 3: Maximum Likelihood Estimation 1/23

Statistics for Applications. Chapter 3: Maximum Likelihood Estimation 1/23 18.650 Statistics for Applicatios Chapter 3: Maximum Likelihood Estimatio 1/23 Total variatio distace (1) ( ) Let E,(IPθ ) θ Θ be a statistical model associated with a sample of i.i.d. r.v. X 1,...,X.

More information

Entropy Rates and Asymptotic Equipartition

Entropy Rates and Asymptotic Equipartition Chapter 29 Etropy Rates ad Asymptotic Equipartitio Sectio 29. itroduces the etropy rate the asymptotic etropy per time-step of a stochastic process ad shows that it is well-defied; ad similarly for iformatio,

More information

MATH 31B: MIDTERM 2 REVIEW

MATH 31B: MIDTERM 2 REVIEW MATH 3B: MIDTERM REVIEW JOE HUGHES. Evaluate x (x ) (x 3).. Partial Fractios Solutio: The umerator has degree less tha the deomiator, so we ca use partial fractios. Write x (x ) (x 3) = A x + A (x ) +

More information

Lecture 19: Convergence

Lecture 19: Convergence Lecture 19: Covergece Asymptotic approach I statistical aalysis or iferece, a key to the success of fidig a good procedure is beig able to fid some momets ad/or distributios of various statistics. I may

More information

Direction: This test is worth 250 points. You are required to complete this test within 50 minutes.

Direction: This test is worth 250 points. You are required to complete this test within 50 minutes. Term Test October 3, 003 Name Math 56 Studet Number Directio: This test is worth 50 poits. You are required to complete this test withi 50 miutes. I order to receive full credit, aswer each problem completely

More information

CONVERGENCE TO FISHER INFORMATION AND THE CENTRAL LIMIT THEOREM

CONVERGENCE TO FISHER INFORMATION AND THE CENTRAL LIMIT THEOREM Techical Scieces ad Applied Matheatics CONVGNC TO FISH INFOMATION AND TH CNTAL LIMIT THOM Bogda Gheorghe MUNTANU, Doru LUCULSCU Heri Coadă Air Forces Acadey, Braşov Abstract: I this paper, we give coditios

More information

EFFECTIVE WLLN, SLLN, AND CLT IN STATISTICAL MODELS

EFFECTIVE WLLN, SLLN, AND CLT IN STATISTICAL MODELS EFFECTIVE WLLN, SLLN, AND CLT IN STATISTICAL MODELS Ryszard Zieliński Ist Math Polish Acad Sc POBox 21, 00-956 Warszawa 10, Polad e-mail: rziel@impagovpl ABSTRACT Weak laws of large umbers (W LLN), strog

More information

Chapter 2. Asymptotic Notation

Chapter 2. Asymptotic Notation Asyptotic Notatio 3 Chapter Asyptotic Notatio Goal : To siplify the aalysis of ruig tie by gettig rid of details which ay be affected by specific ipleetatio ad hardware. [1] The Big Oh (O-Notatio) : It

More information

Distribution of Random Samples & Limit theorems

Distribution of Random Samples & Limit theorems STAT/MATH 395 A - PROBABILITY II UW Witer Quarter 2017 Néhémy Lim Distributio of Radom Samples & Limit theorems 1 Distributio of i.i.d. Samples Motivatig example. Assume that the goal of a study is to

More information

Integrals of Functions of Several Variables

Integrals of Functions of Several Variables Itegrals of Fuctios of Several Variables We ofte resort to itegratios i order to deterie the exact value I of soe quatity which we are uable to evaluate by perforig a fiite uber of additio or ultiplicatio

More information

Exponential Functions and Taylor Series

Exponential Functions and Taylor Series Expoetial Fuctios ad Taylor Series James K. Peterso Departmet of Biological Scieces ad Departmet of Mathematical Scieces Clemso Uiversity March 29, 207 Outlie Revistig the Expoetial Fuctio Taylor Series

More information

Integrable Functions. { f n } is called a determining sequence for f. If f is integrable with respect to, then f d does exist as a finite real number

Integrable Functions. { f n } is called a determining sequence for f. If f is integrable with respect to, then f d does exist as a finite real number MATH 532 Itegrable Fuctios Dr. Neal, WKU We ow shall defie what it meas for a measurable fuctio to be itegrable, show that all itegral properties of simple fuctios still hold, ad the give some coditios

More information

Binomial transform of products

Binomial transform of products Jauary 02 207 Bioial trasfor of products Khristo N Boyadzhiev Departet of Matheatics ad Statistics Ohio Norther Uiversity Ada OH 4580 USA -boyadzhiev@ouedu Abstract Give the bioial trasfors { b } ad {

More information

Notes 5 : More on the a.s. convergence of sums

Notes 5 : More on the a.s. convergence of sums Notes 5 : More o the a.s. covergece of sums Math 733-734: Theory of Probability Lecturer: Sebastie Roch Refereces: Dur0, Sectios.5; Wil9, Sectio 4.7, Shi96, Sectio IV.4, Dur0, Sectio.. Radom series. Three-series

More information

This section is optional.

This section is optional. 4 Momet Geeratig Fuctios* This sectio is optioal. The momet geeratig fuctio g : R R of a radom variable X is defied as g(t) = E[e tx ]. Propositio 1. We have g () (0) = E[X ] for = 1, 2,... Proof. Therefore

More information

DISTANCE BETWEEN UNCERTAIN RANDOM VARIABLES

DISTANCE BETWEEN UNCERTAIN RANDOM VARIABLES MATHEMATICAL MODELLING OF ENGINEERING PROBLEMS Vol, No, 4, pp5- http://doiorg/88/ep4 DISTANCE BETWEEN UNCERTAIN RANDOM VARIABLES Yogchao Hou* ad Weicai Peg Departet of Matheatical Scieces, Chaohu Uiversity,

More information

b i u x i U a i j u x i u x j

b i u x i U a i j u x i u x j M ath 5 2 7 Fall 2 0 0 9 L ecture 1 9 N ov. 1 6, 2 0 0 9 ) S ecod- Order Elliptic Equatios: Weak S olutios 1. Defiitios. I this ad the followig two lectures we will study the boudary value problem Here

More information

Large Sample Theory. Convergence. Central Limit Theorems Asymptotic Distribution Delta Method. Convergence in Probability Convergence in Distribution

Large Sample Theory. Convergence. Central Limit Theorems Asymptotic Distribution Delta Method. Convergence in Probability Convergence in Distribution Large Sample Theory Covergece Covergece i Probability Covergece i Distributio Cetral Limit Theorems Asymptotic Distributio Delta Method Covergece i Probability A sequece of radom scalars {z } = (z 1,z,

More information

Some remarks on the paper Some elementary inequalities of G. Bennett

Some remarks on the paper Some elementary inequalities of G. Bennett Soe rears o the paper Soe eleetary iequalities of G. Beett Dag Ah Tua ad Luu Quag Bay Vieta Natioal Uiversity - Haoi Uiversity of Sciece Abstract We give soe couterexaples ad soe rears of soe of the corollaries

More information

Lecture 13: Maximum Likelihood Estimation

Lecture 13: Maximum Likelihood Estimation ECE90 Sprig 007 Statistical Learig Theory Istructor: R. Nowak Lecture 3: Maximum Likelihood Estimatio Summary of Lecture I the last lecture we derived a risk (MSE) boud for regressio problems; i.e., select

More information

6. Sufficient, Complete, and Ancillary Statistics

6. Sufficient, Complete, and Ancillary Statistics Sufficiet, Complete ad Acillary Statistics http://www.math.uah.edu/stat/poit/sufficiet.xhtml 1 of 7 7/16/2009 6:13 AM Virtual Laboratories > 7. Poit Estimatio > 1 2 3 4 5 6 6. Sufficiet, Complete, ad Acillary

More information

s = and t = with C ij = A i B j F. (i) Note that cs = M and so ca i µ(a i ) I E (cs) = = c a i µ(a i ) = ci E (s). (ii) Note that s + t = M and so

s = and t = with C ij = A i B j F. (i) Note that cs = M and so ca i µ(a i ) I E (cs) = = c a i µ(a i ) = ci E (s). (ii) Note that s + t = M and so 3 From the otes we see that the parts of Theorem 4. that cocer us are: Let s ad t be two simple o-egative F-measurable fuctios o X, F, µ ad E, F F. The i I E cs ci E s for all c R, ii I E s + t I E s +

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 19 11/17/2008 LAWS OF LARGE NUMBERS II THE STRONG LAW OF LARGE NUMBERS

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 19 11/17/2008 LAWS OF LARGE NUMBERS II THE STRONG LAW OF LARGE NUMBERS MASSACHUSTTS INSTITUT OF TCHNOLOGY 6.436J/5.085J Fall 2008 Lecture 9 /7/2008 LAWS OF LARG NUMBRS II Cotets. The strog law of large umbers 2. The Cheroff boud TH STRONG LAW OF LARG NUMBRS While the weak

More information

2 Banach spaces and Hilbert spaces

2 Banach spaces and Hilbert spaces 2 Baach spaces ad Hilbert spaces Tryig to do aalysis i the ratioal umbers is difficult for example cosider the set {x Q : x 2 2}. This set is o-empty ad bouded above but does ot have a least upper boud

More information

7.1 Convergence of sequences of random variables

7.1 Convergence of sequences of random variables Chapter 7 Limit Theorems Throughout this sectio we will assume a probability space (, F, P), i which is defied a ifiite sequece of radom variables (X ) ad a radom variable X. The fact that for every ifiite

More information

Automated Proofs for Some Stirling Number Identities

Automated Proofs for Some Stirling Number Identities Autoated Proofs for Soe Stirlig Nuber Idetities Mauel Kauers ad Carste Scheider Research Istitute for Sybolic Coputatio Johaes Kepler Uiversity Altebergerstraße 69 A4040 Liz, Austria Subitted: Sep 1, 2007;

More information

Summary. Recap ... Last Lecture. Summary. Theorem

Summary. Recap ... Last Lecture. Summary. Theorem Last Lecture Biostatistics 602 - Statistical Iferece Lecture 23 Hyu Mi Kag April 11th, 2013 What is p-value? What is the advatage of p-value compared to hypothesis testig procedure with size α? How ca

More information

Maximum Likelihood Estimation

Maximum Likelihood Estimation ECE 645: Estimatio Theory Sprig 2015 Istructor: Prof. Staley H. Cha Maximum Likelihood Estimatio (LaTeX prepared by Shaobo Fag) April 14, 2015 This lecture ote is based o ECE 645(Sprig 2015) by Prof. Staley

More information

Exponential Functions and Taylor Series

Exponential Functions and Taylor Series MATH 4530: Aalysis Oe Expoetial Fuctios ad Taylor Series James K. Peterso Departmet of Biological Scieces ad Departmet of Mathematical Scieces Clemso Uiversity March 29, 2017 MATH 4530: Aalysis Oe Outlie

More information

Advanced Analysis. Min Yan Department of Mathematics Hong Kong University of Science and Technology

Advanced Analysis. Min Yan Department of Mathematics Hong Kong University of Science and Technology Advaced Aalysis Mi Ya Departmet of Mathematics Hog Kog Uiversity of Sciece ad Techology September 3, 009 Cotets Limit ad Cotiuity 7 Limit of Sequece 8 Defiitio 8 Property 3 3 Ifiity ad Ifiitesimal 8 4

More information

Information Theory Tutorial Communication over Channels with memory. Chi Zhang Department of Electrical Engineering University of Notre Dame

Information Theory Tutorial Communication over Channels with memory. Chi Zhang Department of Electrical Engineering University of Notre Dame Iformatio Theory Tutorial Commuicatio over Chaels with memory Chi Zhag Departmet of Electrical Egieerig Uiversity of Notre Dame Abstract A geeral capacity formula C = sup I(; Y ), which is correct for

More information

Random Walks on Discrete and Continuous Circles. by Jeffrey S. Rosenthal School of Mathematics, University of Minnesota, Minneapolis, MN, U.S.A.

Random Walks on Discrete and Continuous Circles. by Jeffrey S. Rosenthal School of Mathematics, University of Minnesota, Minneapolis, MN, U.S.A. Radom Walks o Discrete ad Cotiuous Circles by Jeffrey S. Rosethal School of Mathematics, Uiversity of Miesota, Mieapolis, MN, U.S.A. 55455 (Appeared i Joural of Applied Probability 30 (1993), 780 789.)

More information

Solutions to Tutorial 5 (Week 6)

Solutions to Tutorial 5 (Week 6) The Uiversity of Sydey School of Mathematics ad Statistics Solutios to Tutorial 5 (Wee 6 MATH2962: Real ad Complex Aalysis (Advaced Semester, 207 Web Page: http://www.maths.usyd.edu.au/u/ug/im/math2962/

More information

Sequences III. Chapter Roots

Sequences III. Chapter Roots Chapter 4 Sequeces III 4. Roots We ca use the results we ve established i the last workbook to fid some iterestig limits for sequeces ivolvig roots. We will eed more techical expertise ad low cuig tha

More information

MAT1026 Calculus II Basic Convergence Tests for Series

MAT1026 Calculus II Basic Convergence Tests for Series MAT026 Calculus II Basic Covergece Tests for Series Egi MERMUT 202.03.08 Dokuz Eylül Uiversity Faculty of Sciece Departmet of Mathematics İzmir/TURKEY Cotets Mootoe Covergece Theorem 2 2 Series of Real

More information

Information Theory and Statistics Lecture 4: Lempel-Ziv code

Information Theory and Statistics Lecture 4: Lempel-Ziv code Iformatio Theory ad Statistics Lecture 4: Lempel-Ziv code Łukasz Dębowski ldebowsk@ipipa.waw.pl Ph. D. Programme 203/204 Etropy rate is the limitig compressio rate Theorem For a statioary process (X i)

More information

4.1 Data processing inequality

4.1 Data processing inequality ECE598: Iformatio-theoretic methods i high-dimesioal statistics Sprig 206 Lecture 4: Total variatio/iequalities betwee f-divergeces Lecturer: Yihog Wu Scribe: Matthew Tsao, Feb 8, 206 [Ed. Mar 22] Recall

More information

Online Learning & Game Theory

Online Learning & Game Theory Olie Learig & Gae Theory A quick overview with recet results Viaey Perchet Laboratoire Probabilités et Modèles Aléatoires Uiv. Paris-Diderot Jourées MAS 2014 27 Août 2014 Startig Exaples Startig Exaples

More information

Product measures, Tonelli s and Fubini s theorems For use in MAT3400/4400, autumn 2014 Nadia S. Larsen. Version of 13 October 2014.

Product measures, Tonelli s and Fubini s theorems For use in MAT3400/4400, autumn 2014 Nadia S. Larsen. Version of 13 October 2014. Product measures, Toelli s ad Fubii s theorems For use i MAT3400/4400, autum 2014 Nadia S. Larse Versio of 13 October 2014. 1. Costructio of the product measure The purpose of these otes is to preset the

More information

Power series are analytic

Power series are analytic Power series are aalytic Horia Corea 1 1 The expoetial ad the logarithm For every x R we defie the fuctio give by exp(x) := 1 + x + x + + x + = x. If x = 0 we have exp(0) = 1. If x 0, cosider the series

More information

Entropy and Ergodic Theory Lecture 5: Joint typicality and conditional AEP

Entropy and Ergodic Theory Lecture 5: Joint typicality and conditional AEP Etropy ad Ergodic Theory Lecture 5: Joit typicality ad coditioal AEP 1 Notatio: from RVs back to distributios Let (Ω, F, P) be a probability space, ad let X ad Y be A- ad B-valued discrete RVs, respectively.

More information

Let us give one more example of MLE. Example 3. The uniform distribution U[0, θ] on the interval [0, θ] has p.d.f.

Let us give one more example of MLE. Example 3. The uniform distribution U[0, θ] on the interval [0, θ] has p.d.f. Lecture 5 Let us give oe more example of MLE. Example 3. The uiform distributio U[0, ] o the iterval [0, ] has p.d.f. { 1 f(x =, 0 x, 0, otherwise The likelihood fuctio ϕ( = f(x i = 1 I(X 1,..., X [0,

More information

Introduction to Extreme Value Theory Laurens de Haan, ISM Japan, Erasmus University Rotterdam, NL University of Lisbon, PT

Introduction to Extreme Value Theory Laurens de Haan, ISM Japan, Erasmus University Rotterdam, NL University of Lisbon, PT Itroductio to Extreme Value Theory Laures de Haa, ISM Japa, 202 Itroductio to Extreme Value Theory Laures de Haa Erasmus Uiversity Rotterdam, NL Uiversity of Lisbo, PT Itroductio to Extreme Value Theory

More information

Week 10 Spring Lecture 19. Estimation of Large Covariance Matrices: Upper bound Observe. is contained in the following parameter space,

Week 10 Spring Lecture 19. Estimation of Large Covariance Matrices: Upper bound Observe. is contained in the following parameter space, Week 0 Sprig 009 Lecture 9. stiatio of Large Covariace Matrices: Upper boud Observe ; ; : : : ; i.i.d. fro a p-variate Gaussia distributio, N (; pp ). We assue that the covariace atrix pp = ( ij ) i;jp

More information

Limit Theorems. Convergence in Probability. Let X be the number of heads observed in n tosses. Then, E[X] = np and Var[X] = np(1-p).

Limit Theorems. Convergence in Probability. Let X be the number of heads observed in n tosses. Then, E[X] = np and Var[X] = np(1-p). Limit Theorems Covergece i Probability Let X be the umber of heads observed i tosses. The, E[X] = p ad Var[X] = p(-p). L O This P x p NM QP P x p should be close to uity for large if our ituitio is correct.

More information

SUMMARY OF SEQUENCES AND SERIES

SUMMARY OF SEQUENCES AND SERIES SUMMARY OF SEQUENCES AND SERIES Importat Defiitios, Results ad Theorems for Sequeces ad Series Defiitio. A sequece {a } has a limit L ad we write lim a = L if for every ɛ > 0, there is a correspodig iteger

More information

Chapter 3. Strong convergence. 3.1 Definition of almost sure convergence

Chapter 3. Strong convergence. 3.1 Definition of almost sure convergence Chapter 3 Strog covergece As poited out i the Chapter 2, there are multiple ways to defie the otio of covergece of a sequece of radom variables. That chapter defied covergece i probability, covergece i

More information

Power series are analytic

Power series are analytic Power series are aalytic Horia Corea 1 1 Fubii s theorem for double series Theorem 1.1. Let {α m }, be a real sequece idexed by two idices. Assume that the series α m is coverget for all ad C := ( α m

More information

Moment-entropy inequalities for a random vector

Moment-entropy inequalities for a random vector 1 Momet-etropy iequalities for a radom vector Erwi Lutwak, Deae ag, ad Gaoyog Zhag Abstract The p-th momet matrix is defied for a real radom vector, geeralizig the classical covariace matrix. Sharp iequalities

More information

19.1 The dictionary problem

19.1 The dictionary problem CS125 Lecture 19 Fall 2016 19.1 The dictioary proble Cosider the followig data structural proble, usually called the dictioary proble. We have a set of ites. Each ite is a (key, value pair. Keys are i

More information

Application to Random Graphs

Application to Random Graphs A Applicatio to Radom Graphs Brachig processes have a umber of iterestig ad importat applicatios. We shall cosider oe of the most famous of them, the Erdős-Réyi radom graph theory. 1 Defiitio A.1. Let

More information

Ans: a n = 3 + ( 1) n Determine whether the sequence converges or diverges. If it converges, find the limit.

Ans: a n = 3 + ( 1) n Determine whether the sequence converges or diverges. If it converges, find the limit. . Fid a formula for the term a of the give sequece: {, 3, 9, 7, 8 },... As: a = 3 b. { 4, 9, 36, 45 },... As: a = ( ) ( + ) c. {5,, 5,, 5,, 5,,... } As: a = 3 + ( ) +. Determie whether the sequece coverges

More information

Introductory statistics

Introductory statistics CM9S: Machie Learig for Bioiformatics Lecture - 03/3/06 Itroductory statistics Lecturer: Sriram Sakararama Scribe: Sriram Sakararama We will provide a overview of statistical iferece focussig o the key

More information

Mathematical Methods for Physics and Engineering

Mathematical Methods for Physics and Engineering Mathematical Methods for Physics ad Egieerig Lecture otes Sergei V. Shabaov Departmet of Mathematics, Uiversity of Florida, Gaiesville, FL 326 USA CHAPTER The theory of covergece. Numerical sequeces..

More information

Learning Theory for Conditional Risk Minimization: Supplementary Material

Learning Theory for Conditional Risk Minimization: Supplementary Material Learig Theory for Coditioal Risk Miiizatio: Suppleetary Material Alexader Zii IST Austria azii@istacat Christoph H Lapter IST Austria chl@istacat Proofs Proof of Theore After the applicatio of (6) ad (8)

More information

Direction: This test is worth 150 points. You are required to complete this test within 55 minutes.

Direction: This test is worth 150 points. You are required to complete this test within 55 minutes. Term Test 3 (Part A) November 1, 004 Name Math 6 Studet Number Directio: This test is worth 10 poits. You are required to complete this test withi miutes. I order to receive full credit, aswer each problem

More information

Lecture 3: Convergence of Fourier Series

Lecture 3: Convergence of Fourier Series Lecture 3: Covergece of Fourier Series Himashu Tyagi Let f be a absolutely itegrable fuctio o T : [ π,π], i.e., f L (T). For,,... defie ˆf() f(θ)e i θ dθ. π T The series ˆf()e i θ is called the Fourier

More information

Probability for mathematicians INDEPENDENCE TAU

Probability for mathematicians INDEPENDENCE TAU Probability for mathematicias INDEPENDENCE TAU 2013 28 Cotets 3 Ifiite idepedet sequeces 28 3a Idepedet evets........................ 28 3b Idepedet radom variables.................. 33 3 Ifiite idepedet

More information

We are mainly going to be concerned with power series in x, such as. (x)} converges - that is, lims N n

We are mainly going to be concerned with power series in x, such as. (x)} converges - that is, lims N n Review of Power Series, Power Series Solutios A power series i x - a is a ifiite series of the form c (x a) =c +c (x a)+(x a) +... We also call this a power series cetered at a. Ex. (x+) is cetered at

More information

1 = δ2 (0, ), Y Y n nδ. , T n = Y Y n n. ( U n,k + X ) ( f U n,k + Y ) n 2n f U n,k + θ Y ) 2 E X1 2 X1

1 = δ2 (0, ), Y Y n nδ. , T n = Y Y n n. ( U n,k + X ) ( f U n,k + Y ) n 2n f U n,k + θ Y ) 2 E X1 2 X1 8. The cetral limit theorems 8.1. The cetral limit theorem for i.i.d. sequeces. ecall that C ( is N -separatig. Theorem 8.1. Let X 1, X,... be i.i.d. radom variables with EX 1 = ad EX 1 = σ (,. Suppose

More information

EE 4TM4: Digital Communications II Probability Theory

EE 4TM4: Digital Communications II Probability Theory 1 EE 4TM4: Digital Commuicatios II Probability Theory I. RANDOM VARIABLES A radom variable is a real-valued fuctio defied o the sample space. Example: Suppose that our experimet cosists of tossig two fair

More information

McGill University Math 354: Honors Analysis 3 Fall 2012 Solutions to selected problems

McGill University Math 354: Honors Analysis 3 Fall 2012 Solutions to selected problems McGill Uiversity Math 354: Hoors Aalysis 3 Fall 212 Assigmet 3 Solutios to selected problems Problem 1. Lipschitz fuctios. Let Lip K be the set of all fuctios cotiuous fuctios o [, 1] satisfyig a Lipschitz

More information

arxiv: v1 [math.st] 12 Dec 2018

arxiv: v1 [math.st] 12 Dec 2018 DIVERGENCE MEASURES ESTIMATION AND ITS ASYMPTOTIC NORMALITY THEORY : DISCRETE CASE arxiv:181.04795v1 [ath.st] 1 Dec 018 Abstract. 1) BA AMADOU DIADIÉ AND 1,,4) LO GANE SAMB 1. Itroductio 1.1. Motivatios.

More information

Singular Continuous Measures by Michael Pejic 5/14/10

Singular Continuous Measures by Michael Pejic 5/14/10 Sigular Cotiuous Measures by Michael Peic 5/4/0 Prelimiaries Give a set X, a σ-algebra o X is a collectio of subsets of X that cotais X ad ad is closed uder complemetatio ad coutable uios hece, coutable

More information

On the behavior at infinity of an integrable function

On the behavior at infinity of an integrable function O the behavior at ifiity of a itegrable fuctio Emmauel Lesige To cite this versio: Emmauel Lesige. O the behavior at ifiity of a itegrable fuctio. The America Mathematical Mothly, 200, 7 (2), pp.75-8.

More information

Enumerative & Asymptotic Combinatorics

Enumerative & Asymptotic Combinatorics C50 Eumerative & Asymptotic Combiatorics Stirlig ad Lagrage Sprig 2003 This sectio of the otes cotais proofs of Stirlig s formula ad the Lagrage Iversio Formula. Stirlig s formula Theorem 1 (Stirlig s

More information

Lecture 16: UMVUE: conditioning on sufficient and complete statistics

Lecture 16: UMVUE: conditioning on sufficient and complete statistics Lecture 16: UMVUE: coditioig o sufficiet ad complete statistics The 2d method of derivig a UMVUE whe a sufficiet ad complete statistic is available Fid a ubiased estimator of ϑ, say U(X. Coditioig o a

More information

MAS111 Convergence and Continuity

MAS111 Convergence and Continuity MAS Covergece ad Cotiuity Key Objectives At the ed of the course, studets should kow the followig topics ad be able to apply the basic priciples ad theorems therei to solvig various problems cocerig covergece

More information

MINIMAX RATES OF CONVERGENCE AND OPTIMALITY OF BAYES FACTOR WAVELET REGRESSION ESTIMATORS UNDER POINTWISE RISKS

MINIMAX RATES OF CONVERGENCE AND OPTIMALITY OF BAYES FACTOR WAVELET REGRESSION ESTIMATORS UNDER POINTWISE RISKS Statistica Siica 19 (2009) 1389 1406: Supplemet S1 S17 1 MINIMAX RATES OF ONVERGENE AND OPTIMALITY OF BAYES FATOR WAVELET REGRESSION ESTIMATORS UNDER POINTWISE RISKS Natalia Bochkia ad Theofais Sapatias

More information

On Equivalence of Martingale Tail Bounds and Deterministic Regret Inequalities

On Equivalence of Martingale Tail Bounds and Deterministic Regret Inequalities O Equivalece of Martigale Tail Bouds ad Determiistic Regret Iequalities Sasha Rakhli Departmet of Statistics, The Wharto School Uiversity of Pesylvaia Dec 16, 2015 Joit work with K. Sridhara arxiv:1510.03925

More information

Probability and Statistics

Probability and Statistics ICME Refresher Course: robability ad Statistics Staford Uiversity robability ad Statistics Luyag Che September 20, 2016 1 Basic robability Theory 11 robability Spaces A probability space is a triple (Ω,

More information

Some Basic Probability Concepts. 2.1 Experiments, Outcomes and Random Variables

Some Basic Probability Concepts. 2.1 Experiments, Outcomes and Random Variables Some Basic Probability Cocepts 2. Experimets, Outcomes ad Radom Variables A radom variable is a variable whose value is ukow util it is observed. The value of a radom variable results from a experimet;

More information

Element sampling: Part 2

Element sampling: Part 2 Chapter 4 Elemet samplig: Part 2 4.1 Itroductio We ow cosider uequal probability samplig desigs which is very popular i practice. I the uequal probability samplig, we ca improve the efficiecy of the resultig

More information