( θ. sup θ Θ f X (x θ) = L. sup Pr (Λ (X) < c) = α. x : Λ (x) = sup θ H 0. sup θ Θ f X (x θ) = ) < c. NH : θ 1 = θ 2 against AH : θ 1 θ 2
|
|
- Wendy Bryant
- 5 years ago
- Views:
Transcription
1 82 CHAPTER 4. MAXIMUM IKEIHOOD ESTIMATION Defiitio: et X be a radom sample with joit p.m/d.f. f X x θ. The geeralised likelihood ratio test g.l.r.t. of the NH : θ H 0 agaist the alterative AH : θ H 1, give the observed value x, rejects the NH at α -level of sigificace wheever θ Λ x sup θ H 0 f X x θ sup θ Θ f X x θ < c where c is such that Thus the g.l.r.t. has rejectio regio R sup Pr Λ X < c α θ H 0 x : Λ x sup θ H 0 f X x θ sup θ Θ f X x θ θ < c Example: Cosider the last example agai i which we wish to test whether productio i two plats is uiform i.e. test NH : θ 1 θ 2 agaist AH : θ 1 θ 2 where θ i is the parameter of the expoetial distributio of the compoet failure times produced i plat i, i 1, 2. Fid the g.l.r.t. procedure i this case. Solutio: We have see that the likelihood is θ f X x θ θ 1e θ 1T θ m 2 e θ 2S where T z i ad S m y i, ad the log-likelihood is l θ log θ 1 θ 1 T + log θ 2 θ 2 S Hece l θ θ 1 0 l θ θ2 0 θ 1 T 0 m θ 2 S 0 θ1 T θ 2 m S
2 4.4. RESTRICTED M..E AND THE G..R.T. 83 i.e. θ /T m/s ad e m m T T e m S S T S m m T S m e +m We have already see that whe the NH is true the restricted m.l.e. s are θ 1 θ 2 +m ad hece T+S m + m θ e +m T+S + m T e +m T+S S T + S T + S +m + m e +m T + S The g.l.r.t. statistic is therefore Λ x θ +m + m e +m T + S m m T S m e +m ad the rejectio regio of the g.l.r.t. is T S m costat T + S +m R {x :Λ x < c} { } T S m x :costat T + S +m < c { } T S m x : +m < c T + S { T/S } x : +m < c 1 + T/S { } U x : +m < c 1 + U where U T S z i m y i
3 84 CHAPTER 4. MAXIMUM IKEIHOOD ESTIMATION Plot of the statistic U /1+U +m U /1+U +m c 0 0 u 1 u 2 U Figure 4.8: The coditio U /1 + U +m < c is equivalet to U < u 1 ad U > u 2. From the figure above we see that the rejectio regio ca also be writte as R {x :U < u 1 or U > u 2 } { x :U m < v 1 or U m } > v 2 {x :V < v 1 or V > v 2 } where V U m T/ S/m z average of the z-values ȳ average of the y-values. To fid v 1 ad v 2 we must kow the distributio of V whe the NH is true. Sice the Z i s are expoetially distributed with parameter θ 1 the T Z i Gamma,θ 1 i.e. 2θ 1 T χ 2 2. Similarly 2θ 2 S χ 2 2m. Sice the Z i s ad Y i s are idepedet 2θ 1 T/ 2θ 2 S/m χ2 2/2 χ 2 2m/2m F 2,2m Whe the NH is true θ 1 θ 2 ad hece V T/ S/m 2θ 1T/ 2θ 2 S/m F 2,2m
4 4.4. RESTRICTED M..E AND THE G..R.T. 85 The values v 1 ad v 2 determiig the rejectio regio ca therefore be obtaied from the F 2,2m tables such that Pr v 1 F 2,2m v 2 1 α Example: et X be a radom sample of idepedet observatio from the Bi m,θ distributio. Fid the g.l.r.t. of NH : θ θ agaist the alterative AH : θ > θ for some fixed ad kow value θ. Solutio: The likelihood is θ;x where t x i. Cosequetly ad Thus sup θ;x θ Θ sup θ;x θ H 0 Λ x Λ t m x i θ x i 1 θ m x i θ t 1 θ m t sup θ;x θ H 0 sup θ 0,1 θ t 1 θ m t t t 1 t m t m m sup θ t 1 θ m t θ θ t t m 1 t m t t if m m θ θ t 1 θ m t if t m > θ θ;x t 1 if m θ θ t 1 θ m t sup t t θ Θ m 1 t m 1 if t mθ mθ t t m mθ m t m t if t > mθ m t if t m > θ Figure 4.9 gives the plot of Λ t from which it ca be see that it is a o-decreasig fuctio of t ad cosequetly the rejectio regio is give by R {x : Λ t < c α } {x : t > k α }
5 86 CHAPTER 4. MAXIMUM IKEIHOOD ESTIMATION ikelihood fuctio G.l.r.t. statistic 1 likelihood θ g.l.r.t. Λt c α 0 0 t/m 1 θ 0 0 mθ t k α m Figure 4.9: The likelihood of the biomial data ad the g.l.r.t statistic where k α is such that sup Pr X R sup Pr X i > k α α level of sigificace θ θ θ θ But X i Bi m,θ. Further, sice Pr Y > k θ is icreasig i θ whe Y Bi r, θ, it follows that ad k α satisfies sup Pr θ θ m jk a+1 X i > k α Pr m j θ j 1 θ m j α X i > k α θ θ Note that there may ot exist iteger k α which satisfies the above equatio. If it does ot, we the choose k α such that m jk a+1 m j θ j 1 θ m j < α ad m jk a m j θ j 1 θ m j > α
6 4.5. ASYMPTOTIC FORM OF THE G..R.T. 87 i.e. k α is the smallest iteger for which the is less tha α probability that a Bi m,θ radom variable will exceed it. If is large, we ca use the Normal approximatio to the Bi m,θ distributio ad hece obtai k α as the solutio to Pr N mθ,mθ 1 θ > k α α i.e. i.e. Pr N 0, 1 > k α mθ mθ 1 θ α k α mθ mθ 1 θ z α or k α mθ + z α mθ 1 θ where z α is such that Φ z α 1 α, Φ. beig the Stadard Normal distributio fuctio. Cosequetly the rejectio regio is { } R x : x i mθ mθ 1 θ > z α 4.5 Asymptotic form of the g.l.r.t. Result: I testig the NH : θ H 0 agaist the alterative AH ; θ Θ H 0 where H 0 {θ : h 1 θ 0,h 2 θ 0,...,h r θ 0}, provided the sample size o which the test is based is large, the uder mild regularity coditios 2 log Λ X 2 log θ 2[l θ l] is approximately chi-squared distributed with r degrees of freedom whe the ull hypothesis NH is true. The critical regio of the g.l.r.t ca therefore be take as R { } x : 2 log Λ x χ 2 r,α where χ 2 r,α is take from the chi-squared tables ad is such that if W χ 2 r the Pr W χ 2 r,α α ad α level of sigificace of test. The degrees
7 88 CHAPTER 4. MAXIMUM IKEIHOOD ESTIMATION of freedom are equal to the umber of idepedet side coditios used to specify the ull hypothesis. Note that 2 log Λ x χ 2 r idepedetly of θ as log as θ H 0. I the above Λ x is the g.l.r.t. statistic, θ is the likelihood fuctio, lθ is the lo-likelihood fuctio, θ the m.l.e. of θ ad θ the restricted m.l.e. i H 0. Sketch of proof: Sice θ Λ x sup θ θ H 0 sup θ θ Θ we have that { } 2 log Λ x 2 log log θ { } 2 l l θ where θ is the m.l.e of θ ad θ is the restricted m.l.e. However, we have see that whe bmθ ad θ emerge as turig poits of the likelihood θ ad hece of the log-likelihood l θ ad whe the ull hypothesis is true, θ ad θ are close to each other with high probability. Hece expadig l θ about θ we get to a secod order approximatio k 2 log Λ x 2 l l θi θ l i θ i 1 k k θi 2 θ i θj θ 2 l j θ j1 i θ j k k θi θ i θj θ 2 l j 4.19 θ i θ j j1 By argumets similar to those that produced 4.8 we have by the law of large umbers 2 l θ 2 l θ E I ij θ as θ i θ j θ i θ j
8 4.5. ASYMPTOTIC FORM OF THE G..R.T. 89 with I ij θ the i,jth elemet of the Fisher Iformatio matrix I θ. Hece for large 4.19 is further approximated by k k 2 log Λ x θi θ i θj θ j I ij j1 T θ θ I θ θ 4.20 Fially sice θ ad θ 0 are close to each other with high probability we ca itroduce the further approximatio T 2 log Λ x θ θ I θ0 θ θ But sice, as we have see, whe the ull hypothesis is true both θ ad θ are multivariate Normally distributed ad hece so is θ θ it follows that 2 log Λ x is approximately a quadratic i a Normally distributed radom vector of zero mea vector. It ca be show that such a quadratic is chisquared distributed. This is a extesio of the result that if W N 0,σ 2 the W 2 /σ 2 χ 2 1 Example et X ij, j 1, 2,..., i be the failure times of a radom sample of i electroic compoet produced selected from the productio lie of the ith maufacturer i 1, 2, 3 ad assume that they are idepedetly ad expoetially distributed with meas which may differ from maufacturer to maufacturer. Assumig that 1, 2 ad 3 are large, costruct a approximate test to test the ull hypothesis that the mea times to failure are the same for the three compaies. If ad X 1j 3106, X 2j 5620 X 3j 3912 j j carry out the test ad report your coclusios. Solutio: {X ij } i j1 is a radom sample from the expoetial distributio with parameter θ i, mea 1/θ i i 1, 2, 3 ad we wish to test the hypothesis NH : θ 1 θ 2 θ 3 agaist AH : θ 1 θ 2 θ 3 Here θ θ 1,θ 2,θ 3 ad the likelihood of the observed values is θ 1 j1 2 3 θ 1 e θx 1j θ 2 e θx 2j θ 3 e θx 3j j1 j j1 θ 1 1 e θ 1t 1 θ 2 2 e θ 2t 2 θ 3 3 e θ 3t 3
9 90 CHAPTER 4. MAXIMUM IKEIHOOD ESTIMATION where t 1 1 j1 x 1j, t 2 2 j1 x 2j t 3 3 j1 x 3j. The log-likelihood is Hece l θ 1 log θ log θ log θ 3 θ 1 t 1 θ 2 t 2 θ 3 t 3 l θ θ i 0 i θ i t i θ i i t i i 1, 2, 3 Whe NH is true i.e. whe θ 1 θ 2 θ 3 θ say the likelihood fuctio reduces to θ θ e θt 1+t 2 +t 3 ad Hece i.e. Cosequetly l θ log θ log θ θ t 1 + t 2 + t 3 l θ θ θ θ t 1 + t 2 + t 3 0 θ t 1 + t 2 + t 3 θ 1 θ 2 θ e t 1 + t 2 + t 3 ad 1 t 1 1 e 1 2 t 2 2 e e t 1 1 t 2 2 t t 3 3 e 3 Hece g.l.r.t. statistic is θ Λ x t 1 1 t 2 2 t t 1 + t 2 + t t1 t2 t3 t 1 2 3
10 4.5. ASYMPTOTIC FORM OF THE G..R.T. 91 where ad t t 1 + t 2 + t 3. The asymptotic form of the g.l.r.t. statistic is [ ] 2 log Λ x 2 l l θ [ ] t t1 t2 t3 2 log 1 log 2 log 3 log χ 2 2 whe the ull hypothesis is true Note that NH is specified i terms of two coditios sice θ 1 θ 2 θ 3 { θ1 θ 2 0 θ 1 θ 3 0 distributio. For the data give { log Λ 2 60 log 20 log } log 20 log Sice < χ 2 2; there is o evidece, at the 5% level, to reject the ull hypothesis. A importat example: Observatios fall idepedetly i oe of four categories C 1,C 2,c 3 ad C 4 with respective probabilities θ 1,θ 2,θ 3,θ 4 with θ 1 + θ 2 + θ 3 + θ 4 1. the followig hypothesis is put forward NH : θ 1 β 2, θ β1 β, θ β1 β, θ 4 1 β 2 with β 0.6 I a radom sample of 100 such observatios the umbers fallig i the four categories were x 1 10, x 2 13, x 3 37 ad x 4 40 Perform a test of approximate 5% level of sigificace to test this hypothesis agaist the alterative that at least oe of the equalities i NH does ot hold; show that there is evidece to reject NH.
11 92 CHAPTER 4. MAXIMUM IKEIHOOD ESTIMATION Solutio: Note that NH really says θ , θ , θ , θ Note also that sice the θ i s add up to 1, the last oe is determied as soo as we kow the first three θ i s; cosequetly NH ivolves oly 3 idepedet equatios/coditios. The likelihood of θ θ 1,θ 2,θ 3,θ 4 T for the result x i observatios i the ith category i 1, 2, 3, 4 whe are sampled is give by the multiomial probability where x 1 + x 2 + x 3 + x 4 ad! θ x 1!x 2!x 3!x 4! θx 1 1 θ x 2 2 θ x 3 3 θ x 4 4 θ x 1 1 θ x 2 2 θ x θ 1 θ 2 θ 3 x 4 lθ x 1 log θ 1 + x 2 log θ 2 + x 3 log θ 3 + x 4 log θ 1 θ 2 θ 3 + cost] Hece differetiatig w.r.t. θ i i 1, 2, 3 ad equatig to zero to obtai the m.l.e. we get i.e. lθ θ i 0 x i x 4 θ i 1 θ 1 θ 2 θ i 1, 2, 3 3 θ i x i x 4 θ 4 i 1, 2, 3, Sice 4 θ i 1 addig the four equatios above gives 1 x 1 + x 2 + x 3 + x 4 θ4 θ4 x 4 x θ 4 x 4 4 Replacig this i 4.21 we get the m.l.e. θ i x i i 1, 2, 3, 4 ad hece the maximised log-likelihood fuctio l x i log θ i + cost x i log xi + cost
12 4.5. ASYMPTOTIC FORM OF THE G..R.T. 93 Whe NH is true H 0 is a oe poit set, amely H 0 {θ 0.36, 0.12, 0.36, 0.16}. Hece θ 0.36, 0.12, 0.36, 0.16 π 1,π 2, π 3,π 5 ad the maximized log-likelihood uder the NH is l θ x i log π i + cost Therefore [ 2 log Λx 2[l l θ] 2 x i log Note that 2 x i log xi xi π i ] x i log π i χ 2 3 π i expected umber of observatios out of to fall i the ith category whe NH is true e i x i observed umber of observatios out of that fall i the ith category o i Thus the test statistic has the form 2 log Λx 2 o i log oi e i. whe NH is true For the give results 2 log Λx 2 10 log log + 37 log + 40 log > χ 2 3, There is, therefore, evidece at the 5% level of sigificace agaist the ull hypothesis which is therefore rejected.
13 94 CHAPTER 4. MAXIMUM IKEIHOOD ESTIMATION b We ow chage the ull hypothesis to NH : θ 1 β 2, θ β1 β, θ β1 β, θ 4 1 β 2 with β uspecified i 0,1. Notice that this ull hypothesis is specified i terms of oly two idepedet equatios/coditios. As before the equatio for theta 4 is redudat; further, sice β θ 1, the oly idepedet equatios are θ θ1 1 θ 1, θ θ1 1 θ 1 Thus whe the ull hypothesis NH is true the g.l.r.t. statistic 2 log Λx χ 2 2. Now as i part a whe NH is ot true l x i log xi + cost Whe NH is true the likelihood is θ β! x 1!x 2!x 3!x 4! β2x 1 1 β1 β 2 β 2x 1+x 2 +x 3 1 β x 2+x 3 +2x 4 β N 1 β M x2 3 x3 β1 β 1 β 2x 4 2 where N 2x 1 + x 2 + x 3 ad M x 2 + x 3 + 2x 4. Thus whe NH is true the log-likelihood is ad i.e. the m.l.e. β of β satisfies lβ N log β + M log1 β + costat lβ β 0 N β M 1 β 0 N1 β M β 0 β N N + M N 2 For the give data β N
14 4.5. ASYMPTOTIC FORM OF THE G..R.T. 95 The restricted m.l.e. s of the θ i s whe NH is true are θ 1 β 2 π 1 β, θ2 1 2 β1 β π 2 β, θ β1 β π 3 β, θ4 1 β 2 π 4 β Thus l θ x i log θ i + cost x i log π i β + cost so that 2 log Λx 2[l l θ] x i log xi xi x i log π i β x i log π β Note agai that π i β estimated expected umber fallig i the ith category e i Thus the g.l.r.t. statistic still has the form For the give data 2 log Λx 2 o i log oi e , e , 2 e i e , e ad [ 2 log Λx 2 10 log 10 ] log log log < χ 2 2,
15 96 CHAPTER 4. MAXIMUM IKEIHOOD ESTIMATION Thus there is o evidece at the 5% level of sigificace to reject the ull hypothesis. This example ca be geeralized. Result: The multiomial test Suppose each observatio ca fall i oe of k categories C 1,C 2,...,C k with PrC i θ i, i 1, 2,...,k, k θ i 1, ad that observatios are take idepedetly with x i fallig i C i, i 1, 2,...,k, k x i. Thus the sample joit p.m.f. is f X x θ! x 1!x 2!...x k! θx i 1 θ x2 θ x k k We formulate the ull hypothesis that states that the θ i s follow the model NH : θ i π i β i 1, 2,...,k with the π i s give fuctios ivolvig a ukow s-dimesioal parameter β if o β parameter is ivolved i the model the s 0 ad the π is i NH are give values. For large, the asymptotic form of the g.l.r.t. statistic is 2 log Λx k χ 2 k 1 s xi x i log π i β k o i log if s 0 the π i β π i, the give umerical value i NH where β is the m.l.e. of β uder the ull hypothesis i.e β maximises! Q k x i! k [π iβ] x i. oi e i
Goodness-of-Fit Tests and Categorical Data Analysis (Devore Chapter Fourteen)
Goodess-of-Fit Tests ad Categorical Data Aalysis (Devore Chapter Fourtee) MATH-252-01: Probability ad Statistics II Sprig 2019 Cotets 1 Chi-Squared Tests with Kow Probabilities 1 1.1 Chi-Squared Testig................
More informationMath 152. Rumbos Fall Solutions to Review Problems for Exam #2. Number of Heads Frequency
Math 152. Rumbos Fall 2009 1 Solutios to Review Problems for Exam #2 1. I the book Experimetatio ad Measuremet, by W. J. Youde ad published by the by the Natioal Sciece Teachers Associatio i 1962, the
More informationLecture 6 Simple alternatives and the Neyman-Pearson lemma
STATS 00: Itroductio to Statistical Iferece Autum 06 Lecture 6 Simple alteratives ad the Neyma-Pearso lemma Last lecture, we discussed a umber of ways to costruct test statistics for testig a simple ull
More informationDirection: 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 informationLast Lecture. Wald Test
Last Lecture Biostatistics 602 - Statistical Iferece Lecture 22 Hyu Mi Kag April 9th, 2013 Is the exact distributio of LRT statistic typically easy to obtai? How about its asymptotic distributio? For testig
More information5. Likelihood Ratio Tests
1 of 5 7/29/2009 3:16 PM Virtual Laboratories > 9. Hy pothesis Testig > 1 2 3 4 5 6 7 5. Likelihood Ratio Tests Prelimiaries As usual, our startig poit is a radom experimet with a uderlyig sample space,
More informationResampling Methods. X (1/2), i.e., Pr (X i m) = 1/2. We order the data: X (1) X (2) X (n). Define the sample median: ( n.
Jauary 1, 2019 Resamplig Methods Motivatio We have so may estimators with the property θ θ d N 0, σ 2 We ca also write θ a N θ, σ 2 /, where a meas approximately distributed as Oce we have a cosistet estimator
More informationLecture Notes 15 Hypothesis Testing (Chapter 10)
1 Itroductio Lecture Notes 15 Hypothesis Testig Chapter 10) Let X 1,..., X p θ x). Suppose we we wat to kow if θ = θ 0 or ot, where θ 0 is a specific value of θ. For example, if we are flippig a coi, we
More informationStat 319 Theory of Statistics (2) Exercises
Kig Saud Uiversity College of Sciece Statistics ad Operatios Research Departmet Stat 39 Theory of Statistics () Exercises Refereces:. Itroductio to Mathematical Statistics, Sixth Editio, by R. Hogg, J.
More informationEXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY
EXAMINATIONS OF THE ROYAL STATISTICAL SOCIETY GRADUATE DIPLOMA, 016 MODULE : Statistical Iferece Time allowed: Three hours Cadidates should aswer FIVE questios. All questios carry equal marks. The umber
More informationCommon Large/Small Sample Tests 1/55
Commo Large/Small Sample Tests 1/55 Test of Hypothesis for the Mea (σ Kow) Covert sample result ( x) to a z value Hypothesis Tests for µ Cosider the test H :μ = μ H 1 :μ > μ σ Kow (Assume the populatio
More informationTopic 9: Sampling Distributions of Estimators
Topic 9: Samplig Distributios of Estimators Course 003, 2018 Page 0 Samplig distributios of estimators Sice our estimators are statistics (particular fuctios of radom variables), their distributio ca be
More informationSTATISTICAL INFERENCE
STATISTICAL INFERENCE POPULATION AND SAMPLE Populatio = all elemets of iterest Characterized by a distributio F with some parameter θ Sample = the data X 1,..., X, selected subset of the populatio = sample
More informationLecture 7: Properties of Random Samples
Lecture 7: Properties of Radom Samples 1 Cotiued From Last Class Theorem 1.1. Let X 1, X,...X be a radom sample from a populatio with mea µ ad variace σ
More informationx = Pr ( X (n) βx ) =
Exercise 93 / page 45 The desity of a variable X i i 1 is fx α α a For α kow let say equal to α α > fx α α x α Pr X i x < x < Usig a Pivotal Quatity: x α 1 < x < α > x α 1 ad We solve i a similar way as
More informationTMA4245 Statistics. Corrected 30 May and 4 June Norwegian University of Science and Technology Department of Mathematical Sciences.
Norwegia Uiversity of Sciece ad Techology Departmet of Mathematical Scieces Corrected 3 May ad 4 Jue Solutios TMA445 Statistics Saturday 6 May 9: 3: Problem Sow desity a The probability is.9.5 6x x dx
More informationSummary. 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 informationTopic 9: Sampling Distributions of Estimators
Topic 9: Samplig Distributios of Estimators Course 003, 2018 Page 0 Samplig distributios of estimators Sice our estimators are statistics (particular fuctios of radom variables), their distributio ca be
More informationProblem Set 4 Due Oct, 12
EE226: Radom Processes i Systems Lecturer: Jea C. Walrad Problem Set 4 Due Oct, 12 Fall 06 GSI: Assae Gueye This problem set essetially reviews detectio theory ad hypothesis testig ad some basic otios
More informationDirection: 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 informationTopic 18: Composite Hypotheses
Toc 18: November, 211 Simple hypotheses limit us to a decisio betwee oe of two possible states of ature. This limitatio does ot allow us, uder the procedures of hypothesis testig to address the basic questio:
More informationChapter 13: Tests of Hypothesis Section 13.1 Introduction
Chapter 13: Tests of Hypothesis Sectio 13.1 Itroductio RECAP: Chapter 1 discussed the Likelihood Ratio Method as a geeral approach to fid good test procedures. Testig for the Normal Mea Example, discussed
More informationEcon 325 Notes on Point Estimator and Confidence Interval 1 By Hiro Kasahara
Poit Estimator Eco 325 Notes o Poit Estimator ad Cofidece Iterval 1 By Hiro Kasahara Parameter, Estimator, ad Estimate The ormal probability desity fuctio is fully characterized by two costats: populatio
More informationExpectation and Variance of a random variable
Chapter 11 Expectatio ad Variace of a radom variable The aim of this lecture is to defie ad itroduce mathematical Expectatio ad variace of a fuctio of discrete & cotiuous radom variables ad the distributio
More informationEECS564 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 informationTable 12.1: Contingency table. Feature b. 1 N 11 N 12 N 1b 2 N 21 N 22 N 2b. ... a N a1 N a2 N ab
Sectio 12 Tests of idepedece ad homogeeity I this lecture we will cosider a situatio whe our observatios are classified by two differet features ad we would like to test if these features are idepedet
More informationTopic 9: Sampling Distributions of Estimators
Topic 9: Samplig Distributios of Estimators Course 003, 2016 Page 0 Samplig distributios of estimators Sice our estimators are statistics (particular fuctios of radom variables), their distributio ca be
More informationChi-Squared Tests Math 6070, Spring 2006
Chi-Squared Tests Math 6070, Sprig 2006 Davar Khoshevisa Uiversity of Utah February XXX, 2006 Cotets MLE for Goodess-of Fit 2 2 The Multiomial Distributio 3 3 Applicatio to Goodess-of-Fit 6 3 Testig for
More informationProperties and Hypothesis Testing
Chapter 3 Properties ad Hypothesis Testig 3.1 Types of data The regressio techiques developed i previous chapters ca be applied to three differet kids of data. 1. Cross-sectioal data. 2. Time series data.
More information1.010 Uncertainty in Engineering Fall 2008
MIT OpeCourseWare http://ocw.mit.edu.00 Ucertaity i Egieerig Fall 2008 For iformatio about citig these materials or our Terms of Use, visit: http://ocw.mit.edu.terms. .00 - Brief Notes # 9 Poit ad Iterval
More informationStatistical 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 informationIIT 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 informationStatistical 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 informationFinal Examination Statistics 200C. T. Ferguson June 10, 2010
Fial Examiatio Statistics 00C T. Ferguso Jue 0, 00. (a State the Borel-Catelli Lemma ad its coverse. (b Let X,X,... be i.i.d. from a distributio with desity, f(x =θx (θ+ o the iterval (,. For what value
More informationECE 8527: Introduction to Machine Learning and Pattern Recognition Midterm # 1. Vaishali Amin Fall, 2015
ECE 8527: Itroductio to Machie Learig ad Patter Recogitio Midterm # 1 Vaishali Ami Fall, 2015 tue39624@temple.edu Problem No. 1: Cosider a two-class discrete distributio problem: ω 1 :{[0,0], [2,0], [2,2],
More informationStatistical Inference (Chapter 10) Statistical inference = learn about a population based on the information provided by a sample.
Statistical Iferece (Chapter 10) Statistical iferece = lear about a populatio based o the iformatio provided by a sample. Populatio: The set of all values of a radom variable X of iterest. Characterized
More informationIf, for instance, we were required to test whether the population mean μ could be equal to a certain value μ
STATISTICAL INFERENCE INTRODUCTION Statistical iferece is that brach of Statistics i which oe typically makes a statemet about a populatio based upo the results of a sample. I oesample testig, we essetially
More informationIntroductory 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 informationAsymptotics. Hypothesis Testing UMP. Asymptotic Tests and p-values
of the secod half Biostatistics 6 - Statistical Iferece Lecture 6 Fial Exam & Practice Problems for the Fial Hyu Mi Kag Apil 3rd, 3 Hyu Mi Kag Biostatistics 6 - Lecture 6 Apil 3rd, 3 / 3 Rao-Blackwell
More informationThis exam contains 19 pages (including this cover page) and 10 questions. A Formulae sheet is provided with the exam.
Probability ad Statistics FS 07 Secod Sessio Exam 09.0.08 Time Limit: 80 Miutes Name: Studet ID: This exam cotais 9 pages (icludig this cover page) ad 0 questios. A Formulae sheet is provided with the
More informationFACULTY OF MATHEMATICAL STUDIES MATHEMATICS FOR PART I ENGINEERING. Lectures
FACULTY OF MATHEMATICAL STUDIES MATHEMATICS FOR PART I ENGINEERING Lectures MODULE 5 STATISTICS II. Mea ad stadard error of sample data. Biomial distributio. Normal distributio 4. Samplig 5. Cofidece itervals
More informationChapter 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 informationLecture 33: Bootstrap
Lecture 33: ootstrap Motivatio To evaluate ad compare differet estimators, we eed cosistet estimators of variaces or asymptotic variaces of estimators. This is also importat for hypothesis testig ad cofidece
More informationThe variance of a sum of independent variables is the sum of their variances, since covariances are zero. Therefore. V (xi )= n n 2 σ2 = σ2.
SAMPLE STATISTICS A radom sample x 1,x,,x from a distributio f(x) is a set of idepedetly ad idetically variables with x i f(x) for all i Their joit pdf is f(x 1,x,,x )=f(x 1 )f(x ) f(x )= f(x i ) The sample
More informationPower and Type II Error
Statistical Methods I (EXST 7005) Page 57 Power ad Type II Error Sice we do't actually kow the value of the true mea (or we would't be hypothesizig somethig else), we caot kow i practice the type II error
More informationFirst Year Quantitative Comp Exam Spring, Part I - 203A. f X (x) = 0 otherwise
First Year Quatitative Comp Exam Sprig, 2012 Istructio: There are three parts. Aswer every questio i every part. Questio I-1 Part I - 203A A radom variable X is distributed with the margial desity: >
More informationSince X n /n P p, we know that X n (n. Xn (n X n ) Using the asymptotic result above to obtain an approximation for fixed n, we obtain
Assigmet 9 Exercise 5.5 Let X biomial, p, where p 0, 1 is ukow. Obtai cofidece itervals for p i two differet ways: a Sice X / p d N0, p1 p], the variace of the limitig distributio depeds oly o p. Use the
More informationtests 17.1 Simple versus compound
PAS204: Lecture 17. tests UMP ad asymtotic I this lecture, we will idetify UMP tests, wherever they exist, for comarig a simle ull hyothesis with a comoud alterative. We also look at costructig tests based
More informationSTAC51: Categorical data Analysis
STAC51: Categorical data Aalysis Mahida Samarakoo Jauary 28, 2016 Mahida Samarakoo STAC51: Categorical data Aalysis 1 / 35 Table of cotets Iferece for Proportios 1 Iferece for Proportios Mahida Samarakoo
More informationStatistics 3858 : Likelihood Ratio for Multinomial Models
Statistics 3858 : Likelihood Ratio for Multiomial Models Suppose X is multiomial o M categories, that is X Multiomial, p), where p p 1, p 2,..., p M ) A, ad the parameter space is A {p : p j 0, p j 1 }
More informationLecture 11 and 12: Basic estimation theory
Lecture ad 2: Basic estimatio theory Sprig 202 - EE 94 Networked estimatio ad cotrol Prof. Kha March 2 202 I. MAXIMUM-LIKELIHOOD ESTIMATORS The maximum likelihood priciple is deceptively simple. Louis
More information6 Sample Size Calculations
6 Sample Size Calculatios Oe of the major resposibilities of a cliical trial statisticia is to aid the ivestigators i determiig the sample size required to coduct a study The most commo procedure for determiig
More informationKurskod: TAMS11 Provkod: TENB 21 March 2015, 14:00-18:00. English Version (no Swedish Version)
Kurskod: TAMS Provkod: TENB 2 March 205, 4:00-8:00 Examier: Xiagfeg Yag (Tel: 070 2234765). Please aswer i ENGLISH if you ca. a. You are allowed to use: a calculator; formel -och tabellsamlig i matematisk
More informationLet 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[412] A TEST FOR HOMOGENEITY OF THE MARGINAL DISTRIBUTIONS IN A TWO-WAY CLASSIFICATION
[412] A TEST FOR HOMOGENEITY OF THE MARGINAL DISTRIBUTIONS IN A TWO-WAY CLASSIFICATION BY ALAN STUART Divisio of Research Techiques, Lodo School of Ecoomics 1. INTRODUCTION There are several circumstaces
More informationMatrix Representation of Data in Experiment
Matrix Represetatio of Data i Experimet Cosider a very simple model for resposes y ij : y ij i ij, i 1,; j 1,,..., (ote that for simplicity we are assumig the two () groups are of equal sample size ) Y
More informationLecture 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 informationReview Questions, Chapters 8, 9. f(y) = 0, elsewhere. F (y) = f Y(1) = n ( e y/θ) n 1 1 θ e y/θ = n θ e yn
Stat 366 Lab 2 Solutios (September 2, 2006) page TA: Yury Petracheko, CAB 484, yuryp@ualberta.ca, http://www.ualberta.ca/ yuryp/ Review Questios, Chapters 8, 9 8.5 Suppose that Y, Y 2,..., Y deote a radom
More informationData Analysis and Statistical Methods Statistics 651
Data Aalysis ad Statistical Methods Statistics 651 http://www.stat.tamu.edu/~suhasii/teachig.html Suhasii Subba Rao Review of testig: Example The admistrator of a ursig home wats to do a time ad motio
More informationTests of Hypotheses Based on a Single Sample (Devore Chapter Eight)
Tests of Hypotheses Based o a Sigle Sample Devore Chapter Eight MATH-252-01: Probability ad Statistics II Sprig 2018 Cotets 1 Hypothesis Tests illustrated with z-tests 1 1.1 Overview of Hypothesis Testig..........
More informationStat 421-SP2012 Interval Estimation Section
Stat 41-SP01 Iterval Estimatio Sectio 11.1-11. We ow uderstad (Chapter 10) how to fid poit estimators of a ukow parameter. o However, a poit estimate does ot provide ay iformatio about the ucertaity (possible
More informationEfficient GMM LECTURE 12 GMM II
DECEMBER 1 010 LECTURE 1 II Efficiet The estimator depeds o the choice of the weight matrix A. The efficiet estimator is the oe that has the smallest asymptotic variace amog all estimators defied by differet
More informationLast Lecture. Unbiased Test
Last Lecture Biostatistics 6 - Statistical Iferece Lecture Uiformly Most Powerful Test Hyu Mi Kag March 8th, 3 What are the typical steps for costructig a likelihood ratio test? Is LRT statistic based
More informationUniversity of California, Los Angeles Department of Statistics. Hypothesis testing
Uiversity of Califoria, Los Ageles Departmet of Statistics Statistics 100B Elemets of a hypothesis test: Hypothesis testig Istructor: Nicolas Christou 1. Null hypothesis, H 0 (claim about µ, p, σ 2, µ
More informationJanuary 25, 2017 INTRODUCTION TO MATHEMATICAL STATISTICS
Jauary 25, 207 INTRODUCTION TO MATHEMATICAL STATISTICS Abstract. A basic itroductio to statistics assumig kowledge of probability theory.. Probability I a typical udergraduate problem i probability, we
More informationHYPOTHESIS TESTS FOR ONE POPULATION MEAN WORKSHEET MTH 1210, FALL 2018
HYPOTHESIS TESTS FOR ONE POPULATION MEAN WORKSHEET MTH 1210, FALL 2018 We are resposible for 2 types of hypothesis tests that produce ifereces about the ukow populatio mea, µ, each of which has 3 possible
More informationStatistical inference: example 1. Inferential Statistics
Statistical iferece: example 1 Iferetial Statistics POPULATION SAMPLE A clothig store chai regularly buys from a supplier large quatities of a certai piece of clothig. Each item ca be classified either
More information1 of 7 7/16/2009 6:06 AM Virtual Laboratories > 6. Radom Samples > 1 2 3 4 5 6 7 6. Order Statistics Defiitios Suppose agai that we have a basic radom experimet, ad that X is a real-valued radom variable
More informationTAMS24: Notations and Formulas
TAMS4: Notatios ad Formulas Basic otatios ad defiitios X: radom variable stokastiska variabel Mea Vätevärde: µ = X = by Xiagfeg Yag kpx k, if X is discrete, xf Xxdx, if X is cotiuous Variace Varias: =
More informationLecture 2: Monte Carlo Simulation
STAT/Q SCI 43: Itroductio to Resamplig ethods Sprig 27 Istructor: Ye-Chi Che Lecture 2: ote Carlo Simulatio 2 ote Carlo Itegratio Assume we wat to evaluate the followig itegratio: e x3 dx What ca we do?
More informationBayesian Methods: Introduction to Multi-parameter Models
Bayesia Methods: Itroductio to Multi-parameter Models Parameter: θ = ( θ, θ) Give Likelihood p(y θ) ad prior p(θ ), the posterior p proportioal to p(y θ) x p(θ ) Margial posterior ( θ, θ y) is Iterested
More informationChapter 4. Fourier Series
Chapter 4. Fourier Series At this poit we are ready to ow cosider the caoical equatios. Cosider, for eample the heat equatio u t = u, < (4.) subject to u(, ) = si, u(, t) = u(, t) =. (4.) Here,
More informationLecture 6 Chi Square Distribution (χ 2 ) and Least Squares Fitting
Lecture 6 Chi Square Distributio (χ ) ad Least Squares Fittig Chi Square Distributio (χ ) Suppose: We have a set of measuremets {x 1, x, x }. We kow the true value of each x i (x t1, x t, x t ). We would
More informationComposite Hypotheses
Composite Hypotheses March 25-27, 28 For a composite hypothesis, the parameter space Θ is divided ito two disjoit regios, Θ ad Θ 1. The test is writte H : Θ versus H 1 : Θ 1 with H is called the ull hypothesis
More informationApril 18, 2017 CONFIDENCE INTERVALS AND HYPOTHESIS TESTING, UNDERGRADUATE MATH 526 STYLE
April 18, 2017 CONFIDENCE INTERVALS AND HYPOTHESIS TESTING, UNDERGRADUATE MATH 526 STYLE TERRY SOO Abstract These otes are adapted from whe I taught Math 526 ad meat to give a quick itroductio to cofidece
More informationTEACHER CERTIFICATION STUDY GUIDE
COMPETENCY 1. ALGEBRA SKILL 1.1 1.1a. ALGEBRAIC STRUCTURES Kow why the real ad complex umbers are each a field, ad that particular rigs are ot fields (e.g., itegers, polyomial rigs, matrix rigs) Algebra
More information3. Z Transform. Recall that the Fourier transform (FT) of a DT signal xn [ ] is ( ) [ ] = In order for the FT to exist in the finite magnitude sense,
3. Z Trasform Referece: Etire Chapter 3 of text. Recall that the Fourier trasform (FT) of a DT sigal x [ ] is ω ( ) [ ] X e = j jω k = xe I order for the FT to exist i the fiite magitude sese, S = x [
More informationPSYCHOLOGICAL RESEARCH (PYC 304-C) Lecture 9
Hypothesis testig PSYCHOLOGICAL RESEARCH (PYC 34-C Lecture 9 Statistical iferece is that brach of Statistics i which oe typically makes a statemet about a populatio based upo the results of a sample. I
More information17. Joint distributions of extreme order statistics Lehmann 5.1; Ferguson 15
17. Joit distributios of extreme order statistics Lehma 5.1; Ferguso 15 I Example 10., we derived the asymptotic distributio of the maximum from a radom sample from a uiform distributio. We did this usig
More information6.3 Testing Series With Positive Terms
6.3. TESTING SERIES WITH POSITIVE TERMS 307 6.3 Testig Series With Positive Terms 6.3. Review of what is kow up to ow I theory, testig a series a i for covergece amouts to fidig the i= sequece of partial
More informationLecture 6 Chi Square Distribution (χ 2 ) and Least Squares Fitting
Lecture 6 Chi Square Distributio (χ ) ad Least Squares Fittig Chi Square Distributio (χ ) Suppose: We have a set of measuremets {x 1, x, x }. We kow the true value of each x i (x t1, x t, x t ). We would
More informationStat410 Probability and Statistics II (F16)
Some Basic Cocepts of Statistical Iferece (Sec 5.) Suppose we have a rv X that has a pdf/pmf deoted by f(x; θ) or p(x; θ), where θ is called the parameter. I previous lectures, we focus o probability problems
More informationLinear regression. Daniel Hsu (COMS 4771) (y i x T i β)2 2πσ. 2 2σ 2. 1 n. (x T i β y i ) 2. 1 ˆβ arg min. β R n d
Liear regressio Daiel Hsu (COMS 477) Maximum likelihood estimatio Oe of the simplest liear regressio models is the followig: (X, Y ),..., (X, Y ), (X, Y ) are iid radom pairs takig values i R d R, ad Y
More informationExercise 4.3 Use the Continuity Theorem to prove the Cramér-Wold Theorem, Theorem. (1) φ a X(1).
Assigmet 7 Exercise 4.3 Use the Cotiuity Theorem to prove the Cramér-Wold Theorem, Theorem 4.12. Hit: a X d a X implies that φ a X (1) φ a X(1). Sketch of solutio: As we poited out i class, the oly tricky
More informationStatistical Inference Based on Extremum Estimators
T. Rotheberg Fall, 2007 Statistical Iferece Based o Extremum Estimators Itroductio Suppose 0, the true value of a p-dimesioal parameter, is kow to lie i some subset S R p : Ofte we choose to estimate 0
More informationof the matrix is =-85, so it is not positive definite. Thus, the first
BOSTON COLLEGE Departmet of Ecoomics EC771: Ecoometrics Sprig 4 Prof. Baum, Ms. Uysal Solutio Key for Problem Set 1 1. Are the followig quadratic forms positive for all values of x? (a) y = x 1 8x 1 x
More information- E < p. ˆ p q ˆ E = q ˆ = 1 - p ˆ = sample proportion of x failures in a sample size of n. where. x n sample proportion. population proportion
1 Chapter 7 ad 8 Review for Exam Chapter 7 Estimates ad Sample Sizes 2 Defiitio Cofidece Iterval (or Iterval Estimate) a rage (or a iterval) of values used to estimate the true value of the populatio parameter
More informationLESSON 20: HYPOTHESIS TESTING
LESSN 20: YPTESIS TESTING utlie ypothesis testig Tests for the mea Tests for the proportio 1 YPTESIS TESTING TE CNTEXT Example 1: supervisor of a productio lie wats to determie if the productio time of
More informationSTAT431 Review. X = n. n )
STAT43 Review I. Results related to ormal distributio Expected value ad variace. (a) E(aXbY) = aex bey, Var(aXbY) = a VarX b VarY provided X ad Y are idepedet. Normal distributios: (a) Z N(, ) (b) X N(µ,
More informationA statistical method to determine sample size to estimate characteristic value of soil parameters
A statistical method to determie sample size to estimate characteristic value of soil parameters Y. Hojo, B. Setiawa 2 ad M. Suzuki 3 Abstract Sample size is a importat factor to be cosidered i determiig
More informationINFINITE SEQUENCES AND SERIES
11 INFINITE SEQUENCES AND SERIES INFINITE SEQUENCES AND SERIES 11.4 The Compariso Tests I this sectio, we will lear: How to fid the value of a series by comparig it with a kow series. COMPARISON TESTS
More informationLecture 18: Sampling distributions
Lecture 18: Samplig distributios I may applicatios, the populatio is oe or several ormal distributios (or approximately). We ow study properties of some importat statistics based o a radom sample from
More informationUnbiased 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 informationMOST PEOPLE WOULD RATHER LIVE WITH A PROBLEM THEY CAN'T SOLVE, THAN ACCEPT A SOLUTION THEY CAN'T UNDERSTAND.
XI-1 (1074) MOST PEOPLE WOULD RATHER LIVE WITH A PROBLEM THEY CAN'T SOLVE, THAN ACCEPT A SOLUTION THEY CAN'T UNDERSTAND. R. E. D. WOOLSEY AND H. S. SWANSON XI-2 (1075) STATISTICAL DECISION MAKING Advaced
More informationRank tests and regression rank scores tests in measurement error models
Rak tests ad regressio rak scores tests i measuremet error models J. Jurečková ad A.K.Md.E. Saleh Charles Uiversity i Prague ad Carleto Uiversity i Ottawa Abstract The rak ad regressio rak score tests
More informationChi-squared tests Math 6070, Spring 2014
Chi-squared tests Math 6070, Sprig 204 Davar Khoshevisa Uiversity of Utah March, 204 Cotets MLE for goodess-of fit 2 2 The Multivariate ormal distributio 3 3 Cetral limit theorems 5 4 Applicatio to goodess-of-fit
More information1036: Probability & Statistics
036: Probability & Statistics Lecture 0 Oe- ad Two-Sample Tests of Hypotheses 0- Statistical Hypotheses Decisio based o experimetal evidece whether Coffee drikig icreases the risk of cacer i humas. A perso
More informationHomework for 2/3. 1. Determine the values of the following quantities: a. t 0.1,15 b. t 0.05,15 c. t 0.1,25 d. t 0.05,40 e. t 0.
Name: ID: Homework for /3. Determie the values of the followig quatities: a. t 0.5 b. t 0.055 c. t 0.5 d. t 0.0540 e. t 0.00540 f. χ 0.0 g. χ 0.0 h. χ 0.00 i. χ 0.0050 j. χ 0.990 a. t 0.5.34 b. t 0.055.753
More information1+x 1 + α+x. x = 2(α x2 ) 1+x
Math 2030 Homework 6 Solutios # [Problem 5] For coveiece we let α lim sup a ad β lim sup b. Without loss of geerality let us assume that α β. If α the by assumptio β < so i this case α + β. By Theorem
More informationTheorem. Assume the following (Cramér) conditions 1. θ θ
ML test i a slightly differet form ML Testig (Likelihood Ratio Testig) for o-gaussia models Surya Tokdar Model X f (x θ), θ Θ. Hypothesist H 0 : θ Θ 0 Good set: B c (x) ={θ : l x (θ) max θ Θ l x (θ) c2
More information