Continuous Random Variables: Conditioning, Expectation and Independence

Similar documents
Continuous Random Variables: Conditioning, Expectation and Independence

ρ < 1 be five real numbers. The

Random Variables. ECE 313 Probability with Engineering Applications Lecture 8 Professor Ravi K. Iyer University of Illinois

Entropy, Relative Entropy and Mutual Information

Probability and Statistics. What is probability? What is statistics?

2. Independence and Bernoulli Trials

Dr. Shalabh Department of Mathematics and Statistics Indian Institute of Technology Kanpur

Chapter 4 Multiple Random Variables

CS 2750 Machine Learning Lecture 5. Density estimation. Density estimation

CHAPTER 3 POSTERIOR DISTRIBUTIONS

UNIVERSITY OF OSLO DEPARTMENT OF ECONOMICS

Special Instructions / Useful Data

Functions of Random Variables

å 1 13 Practice Final Examination Solutions - = CS109 Dec 5, 2018

STK3100 and STK4100 Autumn 2017

Chapter 5 Properties of a Random Sample

Midterm Exam 1, section 2 (Solution) Thursday, February hour, 15 minutes

Lecture 3 Probability review (cont d)

2SLS Estimates ECON In this case, begin with the assumption that E[ i

( ) = ( ) ( ) Chapter 13 Asymptotic Theory and Stochastic Regressors. Stochastic regressors model

CHAPTER 6. d. With success = observation greater than 10, x = # of successes = 4, and

STK3100 and STK4100 Autumn 2018

X X X E[ ] E X E X. is the ()m n where the ( i,)th. j element is the mean of the ( i,)th., then

Set Theory and Probability

Lecture Notes Types of economic variables

Econometric Methods. Review of Estimation

Basics of Information Theory: Markku Juntti. Basic concepts and tools 1 Introduction 2 Entropy, relative entropy and mutual information

Chain Rules for Entropy

ENGI 4421 Joint Probability Distributions Page Joint Probability Distributions [Navidi sections 2.5 and 2.6; Devore sections

Chapter 14 Logistic Regression Models

IS 709/809: Computational Methods in IS Research. Simple Markovian Queueing Model

Continuous Distributions

Point Estimation: definition of estimators

Chapter 5 Properties of a Random Sample

Lecture 9. Some Useful Discrete Distributions. Some Useful Discrete Distributions. The observations generated by different experiments have

best estimate (mean) for X uncertainty or error in the measurement (systematic, random or statistical) best

Channel Models with Memory. Channel Models with Memory. Channel Models with Memory. Channel Models with Memory

Training Sample Model: Given n observations, [[( Yi, x i the sample model can be expressed as (1) where, zero and variance σ

Summary of the lecture in Biostatistics

X ε ) = 0, or equivalently, lim

= lim. (x 1 x 2... x n ) 1 n. = log. x i. = M, n

Section 2 Notes. Elizabeth Stone and Charles Wang. January 15, Expectation and Conditional Expectation of a Random Variable.

UNIVERSITY OF OSLO DEPARTMENT OF ECONOMICS

D KL (P Q) := p i ln p i q i

Multivariate Transformation of Variables and Maximum Likelihood Estimation

Artificial Intelligence Learning of decision trees

Chapter 4 Multiple Random Variables

Unit 9. The Tangent Bundle

CHAPTER VI Statistical Analysis of Experimental Data

1 Solution to Problem 6.40

Midterm Exam 1, section 1 (Solution) Thursday, February hour, 15 minutes

Assignment 7/MATH 247/Winter, 2010 Due: Friday, March 19. Powers of a square matrix

Lecture 7. Confidence Intervals and Hypothesis Tests in the Simple CLR Model

A NEW LOG-NORMAL DISTRIBUTION

STK4011 and STK9011 Autumn 2016

Linear Regression with One Regressor

Parameter, Statistic and Random Samples

Multiple Choice Test. Chapter Adequacy of Models for Regression

Ordinary Least Squares Regression. Simple Regression. Algebra and Assumptions.

. The set of these sums. be a partition of [ ab, ]. Consider the sum f( x) f( x 1)

Law of Large Numbers

Introduction to Computer Design. Standard Forms for Boolean Functions. Sums and Products. Standard Forms for Boolean Functions (cont ) CMPT-150

Mathematics HL and Further mathematics HL Formula booklet

{ }{ ( )} (, ) = ( ) ( ) ( ) Chapter 14 Exercises in Sampling Theory. Exercise 1 (Simple random sampling): Solution:

Test Paper-II. 1. If sin θ + cos θ = m and sec θ + cosec θ = n, then (a) 2n = m (n 2 1) (b) 2m = n (m 2 1) (c) 2n = m (m 2 1) (d) none of these

22 Nonparametric Methods.

Module 7. Lecture 7: Statistical parameter estimation

Maximum Likelihood Estimation

ELEC 6041 LECTURE NOTES WEEK 1 Dr. Amir G. Aghdam Concordia University

THE ROYAL STATISTICAL SOCIETY 2016 EXAMINATIONS SOLUTIONS HIGHER CERTIFICATE MODULE 5

7. Joint Distributions

Mean is only appropriate for interval or ratio scales, not ordinal or nominal.

Class 13,14 June 17, 19, 2015

Chapter 5. Curve fitting

BASIC PRINCIPLES OF STATISTICS

Asymptotic inference for MLEs under the special exponential family with double-truncation. Takeshi Emura( 江村剛志 )

STRONG CONSISTENCY FOR SIMPLE LINEAR EV MODEL WITH v/ -MIXING

Parameter Estimation

Random Variate Generation ENM 307 SIMULATION. Anadolu Üniversitesi, Endüstri Mühendisliği Bölümü. Yrd. Doç. Dr. Gürkan ÖZTÜRK.

THE ROYAL STATISTICAL SOCIETY HIGHER CERTIFICATE

Discrete Mathematics and Probability Theory Fall 2016 Seshia and Walrand DIS 10b

Geometric Modeling

THE ROYAL STATISTICAL SOCIETY GRADUATE DIPLOMA

σ σ r = x i x N Statistics Formulas Sample Mean Population Mean Interquartile Range Population Variance Population Standard Deviation

Study of Correlation using Bayes Approach under bivariate Distributions

ON BIVARIATE GEOMETRIC DISTRIBUTION. K. Jayakumar, D.A. Mundassery 1. INTRODUCTION

Qualifying Exam Statistical Theory Problem Solutions August 2005

A be a probability space. A random vector

A Remark on the Uniform Convergence of Some Sequences of Functions

Simulation Output Analysis

9.1 Introduction to the probit and logit models

CS475 Parallel Programming

ECE 3800 Probabilistic Methods of Signal and System Analysis

9 U-STATISTICS. Eh =(m!) 1 Eh(X (1),..., X (m ) ) i.i.d

Random Variables and Probability Distributions

Introduction to Probability

ECON 482 / WH Hong The Simple Regression Model 1. Definition of the Simple Regression Model

Modified Cosine Similarity Measure between Intuitionistic Fuzzy Sets

Lecture 3 Naïve Bayes, Maximum Entropy and Text Classification COSI 134

Lecture 3. Sampling, sampling distributions, and parameter estimation

Transcription:

Cotuous Radom Varables: Codtog, xectato ad Ideedece Berl Che Deartmet o Comuter cece & Iormato geerg atoal Tawa ormal Uverst Reerece: - D.. Bertsekas, J.. Tstskls, Itroducto to robablt, ectos 3.4-3.5

Codtog DF Gve a vet (/3) The codtoal DF o a cotuous radom varable, gve a evet I caot be descrbed terms o, the codtoal DF s deed as a oegatve ucto satsg ( B ) ( x)dx B ( x) ormalzato roert ( x ) dx robablt-berl Che

Codtog DF Gve a vet (/3) I ca be descrbed terms o ( s a subset o the real le wth ( ) > 0 ), the codtoal DF s deed as a oegatve ucto x satsg ( x ) 0, ad or a subset ( x ) ( ) ( B ) B ( ), x otherwse ( B, ) ( ) I B ( x ) dx ( ) ( x ) dx B remas the same shae as excet that t s scaled alog the vertcal axs ormalzato roert ( x ) dx ( x ) dx robablt-berl Che 3

Codtog DF Gve a vet (3/3),,, I K are dsjot evets wth ( ) > 0 or each, that orm a artto o the samle sace, the ( x ) ( ) ( x ) Vercato o the above total robablt theorem ( x) ( ) ( x ) x Takg the x () t dt ( ) () t ( x) ( ) ( x) dervatve dt o both sdes wth resectve to x robablt-berl Che 4

Codtoal xectato Gve a vet The codtoal exectato o a cotuous radom varable, gve a evet ( > ), s deed b The codtoal exectato o a ucto g also has the orm Total xectato Theorem Where,, K, are dsjot evets wth ( ) > 0 each, that orm a artto o the samle sace [ ] x ( x ) dx [ g( ) ] g( x) ( x) [ g( )] ( ) g( ) ( ) 0 ( ) [ ] dx or robablt-berl Che 5

Illustratve xamle xamle 3.0. Mea ad Varace o a ecewse Costat DF. uose that the radom varable has the ecewse costat DF / 3, 0 x, ( x) / 3, x, 0, otherwse. Dee evet { les the rst terval [0,] } evet { les the secod terval [,] } ( ) / 3dx / 3, ( ) / 3dx / 3 0 Recall that the mea ad secod momet o a uorm radom varable over a terval ( a + b) / ad ( a + ab b )/ 3 [ a, b ] s + [ ] /, [ ] / 3 [ ] 3 /, [ ] 7 / 3 var [ ] ( ) [ ] + ( ) [ ] / 3 / + / 3 3 / [ ] ( ) [ ] + ( ) [ ] / 3 / 3 + / 3 7 / 3 5 / 9 ( ) 5 / 9 ( 7 / 6) / 36 7 / 6 robablt-berl Che 6

Multle Cotuous Radom Varables Two cotuous radom varables ad assocated wth a commo exermet are jotl cotuous ad ca be descrbed terms o a jot DF satsg ((, ) B) ( x ) ) B, s a oegatve ucto ormalzato robablt, ) dxd, ( a, c) mlarl, ca be vewed as the robablt er ut area the vct o ( a, c) Where δ s a small ostve umber,,, dxd ( a a + δ, c c + δ ) a+ δ c+ δ ) dxd ( a c) δ a c,,, robablt-berl Che 7

Illustratve xamle xamle 3.3. Two-Dmesoal Uorm DF. We are told that the jot DF o the radom varables ad s a costat c o a area ad s zero outsde. Fd the value o c ad the margal DFs o ad. or The a area, corresod ), x g s deed uorm ze o area 0, 4 to be (c. xamle, jot ) or ) 4 ( x) ) 4 d 4 or x 3 3 ( x) ),, 3 4 3 d 4 4 d d DF ) o otherwse 3.) or or ( ) ) dx 4 3 3 ( ) ),, 4 3 dx 4 dx dx or 3 ( ) ) 4, dx 4 4 dx robablt-berl Che 8

Codtog oe Radom Varable o other Two cotuous radom varables ad have a jot DF. For a wth ( ) > 0, the codtoal DF o gve that s deed b ( x ), ) ( ) ormalzato roert ( x ) dx The margal, jot ad codtoal DFs are related to each other b the ollowg ormulas ) ( ) ( x ),, ( x) ) d., margalzato robablt-berl Che 9

Illustratve xamles (/) ( ) otce that the codtoal DF x has the same shae as the jot DF, ), because the ormalzg actor ( ) does ot deed o x ( x 3.5), 3.5) ( 3.5).5) (.5),.5 ) (.5) ( ) Fgure 3.7: Vsualzato o the codtoal DF x. Let, have a jot DF whch s uorm o the set. For each xed, we cosder the jot DF alog the slce ad ormalze t so that t tegrates to ( x 3.5) ( x 3.5), c. examle 3.3 / 4 / 4 / 4 / / / 4 / 4 robablt-berl Che 0

Illustratve xamles (/) xamle 3.5. Crcular Uorm DF. Be throws a dart at a crcular target o radus r. We assume that he alwas hts the target, ad that all ots o mact ) are equall lkel, so that the jot DF, ) o the radom varables x ad s uorm What s the margal DF ( ), ), area o the crcle 0,, x + r πr 0, otherwse ( ) ), dx x + x + r dx πr πr r, r πr (otce here that DF r ) otherwse πr ( ) s ot uorm) r r dx s the crcle dx ( x ), ) ( ) πr r πr, x r For each value, x s uorm robablt-berl Che ( ) + r

xectato o a Fucto o Radom Varables I ad are jotl cotuous radom varables, ad g s some ucto, the Z g(, ) s also a radom varable (ca be cotuous or dscrete) The exectato o Z ca be calculated b [ Z ] [ g( )] g ) ( x ),,, dxd Z b I s a lear ucto o ad, e.g., Z a +, the [ Z ] [ a + b ] a[ ] b[ ] + a b Where ad are scalars robablt-berl Che

robablt-berl Che 3 Codtoal xectato The roertes o ucodtoal exectato carr though, wth the obvous modcatos, to codtoal exectato [ ] ( ) dx x x ( ) [ ] ( ) ( ) dx x x g g ( ) [ ] ( ) ( ) dx x x g g,,

Total robablt/xectato Theorems Total robablt Theorem For a evet ad a cotuous radom varable ( ) ( ) ( ) d Total xectato Theorem For a cotuous radom varables ad [ ] [ ] ( ) [ g ( )] g ( ) d [ ] ( ) [ g (, )] g (, ) d [ ] ( ) d robablt-berl Che 4

Ideedece Two cotuous radom varables ad are deedet ) ( x) ( ), or all x,, Or ( x ) ( x) or all x ad all wth ( ), > 0 Or ( x) ( ) or all ad all x wth ( x), > 0 robablt-berl Che 5

More Factors about Ideedece (/) I two cotuous radom varables ad are deedet, the two evets o the orms ad B are deedet { } { } (, B) x B, ( x ) x B ( x) ( ) x ( x) dx B ( )( B), ddx ddx [ ] d [ ] ( ) robablt-berl Che 6

More Factors about Ideedece (/) I two cotuous radom varables ad are deedet, the [ ] [ ] [ ] ( + ) var ( ) var ( ) var + The radom varables ad h are deedet or a uctos g ad h Thereore, g ( ) ( ) [ g( ) h( )] [ g( )] [ h( )] robablt-berl Che 7

Jot CDFs I ad are two (ether cotuous or dscrete) radom varables, ther jot cumulatve dstrbuto ucto (CDF) s deed b F I ad urther have a jot DF, the d ) ( x, ),,, ( x ) ( s t) F,,,, ) x F, x ) I ca be deretated at the ot dsdt F, ) robablt-berl Che 8

Illustratve xamle xamle 3.0. Ver that ad are descrbed b a uorm DF o the ut square, the the jot CDF s gve b F ) ( x, ) x, or 0 x,, ( 0,) (, ) ( 0,0) (,0) F, x ), ), or all ) the ut square robablt-berl Che 9

Recall: the Dscrete Baes Rule Let,, K, be dsjot evets that orm a artto o the samle sace, ad assume that ( ) 0, or all. The, or a evet such that we have B ( B) > 0 ( B) ( ) ( B ) ( B) ( ) ( ) B ( ) ( ) k k B k ( ) ( ) B ( ) ( B ) + L + ( ) ( B ) Multlcato rule Total robablt theorem robablt-berl Che 0

Ierece ad the Cotuous Baes Rule (/) s we have a model o a uderlg but uobserved heomeo, rereseted b a radom varable wth DF, ad we make a os measuremet, whch s modeled terms o a codtoal DF. Oce the exermetal value o s measured, what ormato does ths rovde o the ukow value o? ( x) Measuremet Ierece ( ) x ( x ) ( x ), ) ( ) ( x) ( ) x () t ( t) dt robablt-berl Che

robablt-berl Che Ierece ad the Cotuous Baes Rule (/) I the uobserved heomeo s heretl dscrete Let s a dscrete radom varable o the orm that reresets the deret dscrete robabltes or the uobserved heomeo o terest, ad be the MF o { } ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) () ( ) + + + δ δ δ δ δ Total robablt theorem

Illustratve xamles (/) xamle 3.8. lghtbulb roduced b the Geeral Illumato Coma s kow to have a exoetall dstrbuted letme. However, the coma has bee exerecg qualt cotrol roblems. O a gve da, the arameter Λ λ o the DF o s actuall a radom varable, uorml dstrbuted the terval [, 3 / ]. I we test a lghtbulb ad record ts letme ( ), what ca we sa about the uderlg arameter λ? Λ Λ Λ λ ( λ ) λe, 0, λ > 0 ( λ ), or λ 3 / 0, otherwse ( λ ) 3 / Λ ( λ ) Λ ( λ ) () t ( t) Λ Λ dt Codtoed o Λ λ, has a exoetal dstrbuto wth arameter λ 3 / λe te λ t, or λ 3/ dt robablt-berl Che 3

robablt-berl Che 4 Illustratve xamles (/) xamle 3.9. gal Detecto. bar sgal s trasmtted, ad we are gve that ad. The receved sgal s, where ormal ose wth zero mea ad ut varace, deedet o. What s the robablt that, as a ucto o the observed value o? ( ) ( ) + ( ) ( ) - - s e s s ad, ad or, / π σ Codtoed o, has a ormal dstrbuto wth mea ad ut varace s s ( ) ( ) ( ) ( ) () ( ) () ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) e e e e e e e e e e e e + + + + + + + + / / / / / / π π π

Rectato CTIO 3.4 Codtog o a vet roblems 4, 7, 8 CTIO 3.5 Multle Cotuous Radom Varables roblems 9, 4, 5, 6, 8 robablt-berl Che 5