A Bayesian perspective on estimating mean, variance, and standard-deviation from data

Size: px
Start display at page:

Download "A Bayesian perspective on estimating mean, variance, and standard-deviation from data"

Transcription

1 Brigham Youg Uiversity BYU ScholarsArchive All Faculty Publicatios A Bayesia perspective o estimatig mea, variace, ad stadard-deviatio from data Travis E. Oliphat Follow this ad additioal works at: Part of the Electrical ad Computer Egieerig Commos BYU ScholarsArchive Citatio Oliphat, Travis E., "A Bayesia perspective o estimatig mea, variace, ad stadard-deviatio from data" (006. All Faculty Publicatios This Peer-Reviewed Article is brought to you for free ad ope access by BYU ScholarsArchive. It has bee accepted for iclusio i All Faculty Publicatios by a authorized admiistrator of BYU ScholarsArchive. For more iformatio, please cotact scholarsarchive@byu.edu.

2 A Bayesia perspective o estimatig mea, variace, ad stadard-deviatio from data Travis E. Oliphat December 5, 006 Abstract After reviewig some classical estimators for mea, variace, ad stadard-deviatio ad showig that u-biased estimates are ot usually desirable, a Bayesia perspective is employed to determie what is kow about mea, variace, ad stadard deviatio give oly that a data set i-fact has a commo mea ad variace. Maximum-etropy is used to argue that the likelihood fuctio i this situatio should be the same as if the data were idepedet ad idetically distributed Gaussia. A oiformative prior is derived for the mea ad variace ad Bayes rule is used to compute the posterior Probability Desity Fuctio (PDF of (µ, σ as well as ( µ, σ i terms of the sufficiet statistics x = i xi ad C = i (xi x. From the joit distributio margials are determied. It is show ( that µ x C is distributed as Studet-t with degrees of freedom, σ C geeralized-gamma with c = ad a =, ad σ C is distributed as is distributed as iverted-gamma with a =. It is suggested to report the mea of these distributios as the estimate (or the peak if is too small for the mea to be defied ad a cofidece iterval surroudig the media. Itroductio A stadard cocept ecoutered by ayoe exposed to data is the idea of computig a mea, a variace, ad a stadard deviatio from the data. This paper will explore various approaches to computig estimates of mea, stadard-deviatio, ad variace from samples ad will coclude by recommedig a Bayesia approach to iferece about these values from data. Typically it is assumed that the data are realizatios of a collectio of idepedet, idetically distributed (i.i.d. radom variables. This radom vector is deoted X = X, X,..., X, ad the joit Probability Desity Fuctio (PDF of X is f X (X = f X (x i.. Traditioal mea estimate Commoly, the mea of X is estimated as the sample average: ˆµ = Oe ca the show i a rather satisfyig fashio that i= X i. i= E ˆµ = E X Var ˆµ = Var X. These statemets are typically used to justify this choice of estimator for the mea as ubiased ad cosistet. Sometimes this estimator is further justified by oticig that it is also the Maximum Likelihood (ML estimate for the mea assumig the oise comes from a expoetial family (e.g. Gaussia.

3 . Traditioal variace estimate The ML estimate for variace assumig Gaussia oise is It is sometimes suggested to use istead to esure that ˆσ ML = ˆσ = (X i ˆµ. i= (X i ˆµ i= E ˆσ = ˆσ. We are supposed to believe that this is preferable to a estimator that istead miimizes some other metric such as the mea-squared error (which icludes both bias ad variace. A good discussio of these cocepts will also metio that if the X i are all ormal radom variables, the ˆµ ad ˆσ are idepedet, ˆµ is ormal, ad ( ˆσ /σ is chi-squared with degrees of freedom. Cofidece itervals ca the be determied from these facts i a straightforward way..3 Stadard-deviatio estimates Typically, stadard-deviatio estimates are obtaied usig ˆσ = ˆσ. Typically, little is the said about the ucertaity of this estimate. Ofte, the square-root of the u-biased variace is take with little justificatio other tha coveiece ad despite the fact that ˆσ is geerally ot a u-biased estimate of σ eve whe ˆσ is..4 Outlie of the paper I this paper, the mea-square error of modified classical estimators for the variace ad stadard-deviatio will be compared. The poit of this compariso will be to elucidate which ormalizatio factor gives the smallest error (uder the hypothesis of ormally-distributed data. While istructive, this compariso does ot ed the discussio as it does ot address the questio of whether or ot the ormalizatio costat should be the oly issue i dispute. As a result, the problem will be addressed from a Bayesia perspective. Uder this perspective, I begi with the assumptio that the data has a commo mea ad variace ad use maximum etropy (with a flat prior to assert that the likelihood fuctio is ormal. Usig a flat prior for µ ad a Jeffrey s prior for σ (ad σ, the posterior probability of (µ, σ ad ( µ, σ is derived. From this joit posterior, the posterior probability for µ, σ, ad σ ca be give which leads to simple rules for a estimate ad cofidece iterval calculatios. Comparig various estimators Assumig the X i come from a stadard ormal populatio with mea µ ad variace σ, three estimators for σ ad σ will be compared i terms of the mea-squared error ad bias: the ubiased estimator: ˆσ UB ad ˆσ UB, the maximum-likelihood estimator: ˆσ ML ad ˆσ ML, ad 3 the Miimum Mea-Squared Estimator (MMSE (amog those of a certai class: ˆσ MMSE ad ˆσ MMSE. All three estimators of both quatities are of the form ˆσ = a ˆσ = (X i ˆµ i= a (X i ˆµ. i=

4 ˆθ For both classes of estimators, the bias, E, ad the mea-square error, E (ˆθ θ = E ˆθ θe ˆθ + θ, will be calculated assumig X i comes from a ormal distributio with mea µ ad variace σ. The idetity ( MSE ˆθ E (ˆθ θ = Var ˆθ + E ˆθ θ will be useful i what follows.. Estimators of variace For all three estimators of variace it is kow that uder the hypothesis of ormally distributed data, ˆσ /aσ is χ ad therefore has mea ad variace (. Cosequetly, E ˆσ = aσ ( E ˆσ 4 = a σ 4 ( + ( = a σ 4 ( E (ˆσ σ = E ˆσ 4 σ E ˆσ + σ 4 = σ 4 a ( a ( +. It ca be show that the maximum-likelihood estimator for σ requires a ML =. The ubiased estimator for σ is obviously a UB =. The miimum mea-square error estimator is foud by differetiatig Eq. (?? ad settig the result equal to zero. This procedure results i a MMSE = +. The three estimators ad their performace are summarized i the followig table: ˆσ E ˆσ MSE ˆσ UB (X i ˆµ σ σ 4 i= ML (X i ˆµ ( σ 4 σ i= MMSE (X i ˆµ σ σ + It is ot difficult to show that for > i= + < <, ad therefore i a mea-square sese, the MMSE ad ML estimators are both better tha the ubiased estimator. This example serves to show a geeral property that improved estimators are usually possible i a mea-square sese by usig biased estimators. Figure shows MSE ˆσ /σ 4 ad E ˆσ /σ for the three estimators whe >. 3 Estimators for σ Estimators for σ are ot ofte discussed, but are ofte used ad should, therefore, receive better treatmet. For ormally distributed data, the maximum likelihood estimator for σ is ˆσ ML = ˆσ ML = (X i ˆµ, i 3

5 Normalized Mea Ubiased Maximum Likelihood Miimum MSE Normalized R MSE Ubiased Maximum Likelihood Miimum MSE Figure : Normalized mea, E ˆσ /σ ad ormalized root-mea-square error (R-MSE, MSE ˆσ /σ 4, of several estimators of σ. 4

6 ad thus a ML =. The mea ad mea-square error for all three estimators ca be computed by oticig that ˆσ /σ a is χ (a chi radom variable with degrees of freedom. Because we ca coclude that where E χ = Γ ( Γ ( ( Γ Var χ = Γ ( ( aγ E ˆσ = σ Γ ( = t σ a E (ˆσ σ = Var ˆσ + (E ˆσ σ ( Γ ( ( = aσ ( aσ aγ Γ ( + σ Γ ( = σ a ( at + t = Γ ( Γ (. From these expressios, the ubiased estimator will result if a UB = /t while the miimum measquare estimator ca be foud by differetiatig with respect to a the expressio for mea-square error ad solvig for a. The result is a MMSE = t / (. The followig table summarizes the estimators ad their performace. UB ML MMSE ˆσ E ˆσ MSE ˆσ (X i ˆµ σ σ t t i= (X i ˆµ t σ σ t i= t (X i ˆµ t σ σ t i= It ca be show (or observed from the plot below that t < t < t. Therefore, comparig the estimators o the basis of mea-squared error results i the MMSE ad the ML estimator outperformig the ubiased estimator. Figure shows plots of E ˆσ /σ ad MSE ˆσ /σ to give some idea of the small-sample performace of these differet estimators o ormal data. 4 Bayesia Perspective Give data {x, x, x 3,..., x }, the task is to fid the mea µ, variace σ = v ad stadard-deviatio σ of these data. As stated the problem does t have a solutio. More iformatio is eeded i order to work 5

7 Normalized Mea Ubiased Maximum Likelihood Miimum MSE Normalized R MSE Ubiased Maximum Likelihood Miimum MSE Figure : Normalized mea, E ˆσ /σ ad ormalized root-mea-square error (R-MSE, MSE ˆσ /σ, of several estimators of σ. 6

8 towards a aswer. First, assume that data has a commo mea ad a commo variace. The priciple of maximum etropy ca the be applied uder these costraits (usig a flat igorace prior to choose the distributio f (X µ, σ = (π / σ exp σ (x i µ. which adds the least amout of iformatio to the problem other tha the assumptio of a commo µ ad σ. Notice that we ca use maximum etropy (with a flat igorace distributio so that etropy is f (x log f (x dx to justify the commo assumptio of ormal i.i.d. data. Usig Bayes rule we fid that f (µ, σ X = f (X µ, σ f (µ, σ f (X = D f (X µ, σ f (µ, σ where D is a ormalizig costat. This distributio tells us all the iformatio that is available about µ ad σ give the data X. We ca use this joit-pdf to estimate µ ad/or σ ad to report cofidece i the estimates. 4. Choosig the prior f (µ, σ Cetral to solvig this problem is choosig the prior kowledge for µ ad σ. Because we ca ormalize the radom variables usig Z µ σ to obtai zero-mea, uit variace radom variables, µ is a locatio parameter, ad σ is a scale parameter. Followig Jayes s reasoig, we choose the prior which expresses complete igorace except for the fact that µ is a locatio parameter ad σ is a scale parameter. I other words, we cosider a ew problem with data x which is shifted ad scaled versio of the old data. The prior i both of these case should be the same fuctio. However the prior has adjusted accordig to well-established rules. This defies a expressio that the prior should satisfy: f (µ, σ = af (µ + b, aσ where a > 0 ad b is a arbitrary real umber. The prior that satisfies this trasformatio equatio is the so-called Jeffrey s prior. f (µ, σ = cost σ. This prior is improper i the sese that it is ot ormalizable by itself. However, whe used to fid the posterior a total ormalizatio costat ca be foud. Specifically, f (µ, σ X = D exp σ+ σ (x i µ i ( = D exp σ+ σ x i µ x i + µ i i = D exp (µ x + C σ+ σ / i where x = i x i C = x x = i x i ( x i = (x i x i i 7

9 D = = 0 σ + C π Γ (. ( exp α + C σ dαdσ / This joit posterior PDF tells the whole story about µ ad σ if oly samples costraied to have the same µ ad σ are give. Usig this joit PDF we ca compute ay desired probability. Notice that > or else D 0 which is expressig the fact that with = there is o iformatio about σ whatever. Later, will be eeded the joit posterior PDF of µ ad v = σ which is f (µ, v X = G f (X µ, v f (µ, v where f (µ, v = cost ν so that we are just as uiformed about v as about σ. The, G = G = 4. Margial distributios 0 + v ( exp (µ x + C v/ C π Γ ( = D. The joit distributios provide all of the iformatio available about the parameters of iterest usig the data ad the assumptios. Notice that these distributios oly deped o the data (specifically, the statistics x ad C, ad ca be used easily to compute cofidece itervals. We ca itegrate out oe of the variables ad get just the margial desity fuctio of µ or σ separately. f (µ X = = D exp 0 σ+ Γ ( Γ ( πc (µ x + C σ dσ / + (µ x C / ( so that µ x C is Studet-t distributed with degrees of freedom. We aturally eed > for this distributio to provide iformatio. Whe = we have a improper distributio for µ proportioal to x µ. For other cases we ca deduce: E µ X = x dµdv Var µ X = C 3 arg max f (µ X µ = x. > 3 The margial distributio of σ is f (σ X = D σ + exp α + C σ / dα π = D σ exp C σ σ > 0 C = exp C σ Γ ( σ > 0. σ 8

10 σ Thus, C is geeralized gamma distributed with shape parameters c = ad a =. If =, the distributio reduces to a improper distributio proportioal to /σ (i.e. we have o additioal iformatio about σ other tha what we started with. For other values of we ca fid: Γ ( E σ X = Γ ( C >, Var σ X = ( 3 Γ Γ ( C > 3. arg max σ f (σ X = C. This distributio does ot have a well-defied mea uless > ad it does ot have a well-defied variace uless > 3. Fially, the margial distributio of v = σ is f (v X = = ( C Γ ( G v + exp α + C ν/ exp C v (+/ v dα v > 0 ν Whe =, this also reduces to a improper distributio proportioal to /v. For other values of, a iverted gamma distributio with a =. Useful parameters of this distributio are E σ X = 3 C > 3 Var σ X = arg max f ( σ X = σ C ( 3 ( 5 + C. > 5 Notice that this distributio does ot have a well-defied mea uless > 3 ad does ot have a well-defied variace uless > 5. To illustrate, the posterior probabilities for various umbers of samples, Figures 3, 4, 5 show ormalized plots of the Studet-t, geeralized Gamma, ad iverted gamma distributios for =3, 0, ad 50, correspodig to the mea, stadard-deviatio, ad variace of the data sample. 4.3 Gaussia approximatios The margial posterior distributios for µ, σ, ad σ all approach Normal distributios as. I particular, the posterior distributio for µ approaches a ormal distributio with mea x ad variace C. The posterior distributio for σ approaches a ormal distributio with mea C ad variace C. Fially, the posterior distributio for σ approaches a ormal distributio with mea C ad variace C. 4.4 Joit MAP estimators Joit Maximum A-Posterior (MAP estimators are sometimes useful Because µ X ad σ X (ad similarly µ X ad σ X are ot idepedet, the joit MAP estimator ca produce differet results tha the margial MAP estimators. These estimators miimize the joitly-uiform loss fuctio. To fid the joit estimator we solve ˆµ, ˆσ = arg max f (µ, σ X µ,σ = arg mi log f (µ, σ X µ,σ = arg mi µ,σ ( + log σ + (µ x + C σ / C is 9

11 Studet-t PDF with - DOF (Mea =3 =0 = (t (, (/ f (/ t=( x/c Figure 3: Graph of the posterior PDF for the mea for several values of. The fuctio f µ (t, ν = Γ( ν+ πνγ( ν + x ν ν+ is the PDF of the Studet-t distributio with ν Degrees Of Freedom (DOF. 0

12 Geeralized gamma PDF with c=-, a=(-/ (Std. =3 =0 = (s / (/,, ( / / f (/ s= (/C Figure 4: Graph of the posterior PDF for the stadard-deviatio for several values of. The fuctio f σ (s, c, a = c xca Γ(a exp ( x c x > 0 is the PDF of the geeralized gamma distributio with shape parameters c ad a.

13 Iverse gamma PDF with a=(-/ (Variace =3 =0 = (v /, ( / / (3/ f v= (/C Figure 5: Graph of the posterior PDF for the stadard-deviatio for several values of. The fuctio f σ (s, a = Γ(a x a exp ( x x > 0 is the PDF of the iverse gamma distributio with shape parameter a.

14 ad thus, 0 = (ˆµ x ˆσ / 0 = + ˆσ (ˆµ x + C ˆσ 3. / Solvig these simultaeously gives ˆµ JMAP = x, ˆσ JMAP = + C. The joit estimator for µ ad v ca also be foud i the same way. + ˆµ, ˆv = arg mi log v + (µ x + C. µ,v v/ Differetiatig results i Solvig these simultaeously gives 0 = ˆµ x ˆv/ 0 = + ν (ˆµ x + C v. / ˆµ JMAP = x ˆσ JMAP = + C. While the estimator for the mea is u-iterestig, a wide variety of ormalizatio costats to C show up i this aalysis. Usig these estimates, requires a particular devotio to maximizig f (µ, σ V istead of other estimatio approaches. The most useful approach to uderstadig estimates of µ, σ, ad σ is to determie cofidece itervals which is the subject of the ext sectio. 4.5 Cofidece itervals Oe of the advatages of the Bayesia perspective is that it automatically provides a method to obtai practical cofidece itervals for the estimates. With the probability desity fuctio give, cofidece itervals ca be costructed by fidig a area straddlig the mea (or the peak, or the media with equal areas o either side. Give the ature of the cofidece iterval as a area there is some aesthetic value i choosig the media as the middle value to surroud Geeral case Suppose there is a parameter with probability desity fuctio f (θ ad cumulative distributio fuctio F (θ. How is a cofidece iterval, a, b, costructed about the mea, peak, ad/or media. The iterval should be such that the probability of θ lyig withi the rage is α 00 percet, where α is a give parameter. I other words, the area uder f (θ over the cofidece iterval should be α. Suppose ˆθ is the positio about which it is desired a equal-area cofidece iterval. The, the two ed poits of the iterval ca be calculated from { P a θ ˆθ } = α } P {ˆθ θ b = α. 3

15 These state that (ˆθ F F (a = α (ˆθ F (b F = α so that (ˆθ a = F F α (ˆθ b = F F + α. (ˆθ For the cofidece iterval about the media, F = so that for that importat case. α a = F + α b = F 4.5. Mea For the case of the mea, we have see that the distributio of freedom. As a result, F µ (q = x + C F t (q; µ X x C/( is Studet-t with degrees of where Ft (q; ν is the iverse cumulative distributio fuctio (cdf of the Studet-t distributio with ν degrees of freedom. I additio, the mea, the media, ad the peak are all the same value. Note also that because the Studet-t distributio is symmetric: ( ( Ft q; ν = Ft + q; ν so that ( C + α a = x F t ( C + α b = x + F t ; ; Stadard deviatio For the case of the stadard deviatio, we have see that the distributio of (σ X gamma with c = ad a =. Therefore F σ (q = ( C F q; C is geeralized where F (q; a is the iverse cumulative distributio fuctio (cdf of the geeralized gamma distributio with parameters c = ad a: F (q; a = { Γ a, Γ (a q } / where Γ ( a, Γ (a, y = y. As a result: ( C a = F (F ˆσ ( C b = F (F ˆσ C ; C ; α ; + α ; 4

16 where F (x; a = Γ ( a, x /Γ (a is the cumulative distributio fuctio (cdf of the geeralized gamma with parameter a ad c =. Whe usig the media as the ceter poit, these expressios simplify to: ( C α a = F ;, ( C + α b = F ;. If the peak of the distributio is used as the ceter poit, the ( ( C a = F F ; ( ( C b = F F ; Variace We have see that the distributio of ( σ X C F σ (q = C F I α ; + α ; is iverted gamma with a =. Therefore, ( q; where F I (q; a is the iverse cdf of the iverted gamma distributio with parameter a: where Γ ( a, Γ (a, y = y. Therefore, F I (q; a = Γ a, Γ (a q a = C ( F I (F I b = C F I ˆσ C ; ( (F I ˆσ C ; α ; + α ; where F I (x; a = Γ ( a, x /Γ (a is the cdf of the iverted gamma with parameter a. Agai, if the media is used as the ceter poit, the these expressio simplify to a = C ( α F I ;, b = C ( + α F I ;. If the peak of the margial distributio is used as the ceter poit, the equatios are a = C ( ( F I F I + ; α ; b = C ( ( F I F I + ; + α ; Notice that the cofidece iterval for the variace is the square of the cofidece iterval for the stadard deviatio whe the media of the distributio is used i both cases. Also, care must be take for large α ad small that oe of the argumets to the iverse cdf are egative. Such a situatio, idicates that symmetry aroud the peak is impossible. Therefore, either the media should be used as the middle poit, or the area should be take from 0 to a upper boud, b. 5

17 5 Discussio Havig leared that the miimum mea-square estimator for θ from data X is E θ X, oe might be surprised by the fact that the expected value of the posterior margial distributio i this case does ot result i the same estimator as the miimum mea-square estimator over all classes of estimators of the form ac eve though it has the same form. For example, the classic estimator for ˆσ that gives miimum MSE is 3 but the Bayesia miimum mea-square estimator is C. Why are these differet? The differece comes i the subtle distictio betwee the two estimators. The former fids the value of a that miimizes the itegral (ac σ f (ˆσ dˆσ + C while the latter fids the fuctio of X (which happes to be ac (X that miimizes the itegral (g (X σ f ( X, σ dσ dx. This fial itegral icludes a averagig itegral agaist the o-iformative prior as well as a itegral over the data. These are two etirely differet optimizatio problems ad should ot be expected to provide the same result. It is importat to uderstad the full probability distributio of µ, σ, ad σ especially whe the umber of data-samples is small. For example, the variace of the posterior probability distributio for σ is ot eve defied if 5. As a result, it ca be impossible to just report the mea ad variace. A cofidece iterval (or high-desity regio aroud the media of the distributio is always possible. 6 Summary ad Coclusios I this paper, a study of several estimators for the mea, variace, ad stadard-deviatio of data was preseted. I particular, it was show that the ubiased estimator for variace so commoly used is ot typically a good choice (especially for small because usig + as a divisor rather tha shriks the mea-square error of the estimator. I additio, a fully Bayesia perspective o the problem of estimatig a commo mea ad variace from samples was preseted usig maximum etropy ad o-iformative priors. The results provide the posterior coditioal PDF of the mea, stadard-deviatio, ad variace from which estimates ad cofidece itervals ca be calculated. The results also emphasize the poit that calculatig the Bayesia Mea-Square Error (MSE is ot ecessarily the same as other o-bayesia defiitios of MSE because it ivolves aother averagig itegral over the prior iformatio o the quatity to be estimated. The mea of the coditioal PDF miimizes Bayesia MSE. Table summarizes the results for uderstadig µ, σ, ad σ from data assumed to have a commo mea ad variace. The table requires oly the sufficiet statistics It is also very useful to ote that x = C = i= x i (x i x i= ( µ x C is Studet-t distributed with degrees of freedom, σ C is geeralized gamma distributed with shape parameters c = ad a =, ad σ C is iverted gamma with shape parameter a =. These fial facts ca be used to compute estimates ad cofidece itervals from stadardized tables ad distributios. 6

18 Table : Summary of posterior probability distributios for µ, σ, ad σ. PDF: f ( X Mode E Var Γ( µ / Γ( + (µ x C πc C x x 3 > 3 (C/ σ ( / exp Γ( C σ σ σ > 0 C Γ( C > Γ( 3 Γ ( Γ ( σ (C/ ( / exp Γ( C (σ (+/ σ σ > 0 + C 3 C > 3 C ( 3 ( 5 C > 3 > 5 7 Refereces Jayes, E. T., Probability Theory: The Logic of Sciece, Cambridge Uiversity Press, 003. Jeffreys, Sir Harold, Theory of Probability, 3rd editio, Oxford Uiversity Press, 96. 7

5. Best Unbiased Estimators

5. Best Unbiased Estimators Best Ubiased Estimators http://www.math.uah.edu/stat/poit/ubiased.xhtml 1 of 7 7/16/2009 6:13 AM Virtual Laboratories > 7. Poit Estimatio > 1 2 3 4 5 6 5. Best Ubiased Estimators Basic Theory Cosider agai

More information

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

point estimator a random variable (like P or X) whose values are used to estimate a population parameter Estimatio We have oted that the pollig problem which attempts to estimate the proportio p of Successes i some populatio ad the measuremet problem which attempts to estimate the mea value µ of some quatity

More information

Statistics for Economics & Business

Statistics for Economics & Business Statistics for Ecoomics & Busiess Cofidece Iterval Estimatio Learig Objectives I this chapter, you lear: To costruct ad iterpret cofidece iterval estimates for the mea ad the proportio How to determie

More information

5 Statistical Inference

5 Statistical Inference 5 Statistical Iferece 5.1 Trasitio from Probability Theory to Statistical Iferece 1. We have ow more or less fiished the probability sectio of the course - we ow tur attetio to statistical iferece. I statistical

More information

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

Chapter 8. Confidence Interval Estimation. Copyright 2015, 2012, 2009 Pearson Education, Inc. Chapter 8, Slide 1 Chapter 8 Cofidece Iterval Estimatio Copyright 2015, 2012, 2009 Pearso Educatio, Ic. Chapter 8, Slide 1 Learig Objectives I this chapter, you lear: To costruct ad iterpret cofidece iterval estimates for

More information

14.30 Introduction to Statistical Methods in Economics Spring 2009

14.30 Introduction to Statistical Methods in Economics Spring 2009 MIT OpeCourseWare http://ocwmitedu 430 Itroductio to Statistical Methods i Ecoomics Sprig 009 For iformatio about citig these materials or our Terms of Use, visit: http://ocwmitedu/terms 430 Itroductio

More information

Estimating Proportions with Confidence

Estimating Proportions with Confidence Aoucemets: Discussio today is review for midterm, o credit. You may atted more tha oe discussio sectio. Brig sheets of otes ad calculator to midterm. We will provide Scatro form. Homework: (Due Wed Chapter

More information

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

Lecture 4: Parameter Estimation and Confidence Intervals. GENOME 560 Doug Fowler, GS Lecture 4: Parameter Estimatio ad Cofidece Itervals GENOME 560 Doug Fowler, GS (dfowler@uw.edu) 1 Review: Probability Distributios Discrete: Biomial distributio Hypergeometric distributio Poisso distributio

More information

Bayes Estimator for Coefficient of Variation and Inverse Coefficient of Variation for the Normal Distribution

Bayes Estimator for Coefficient of Variation and Inverse Coefficient of Variation for the Normal Distribution Iteratioal Joural of Statistics ad Systems ISSN 0973-675 Volume, Number 4 (07, pp. 7-73 Research Idia Publicatios http://www.ripublicatio.com Bayes Estimator for Coefficiet of Variatio ad Iverse Coefficiet

More information

x satisfying all regularity conditions. Then

x satisfying all regularity conditions. Then AMS570.01 Practice Midterm Exam Sprig, 018 Name: ID: Sigature: Istructio: This is a close book exam. You are allowed oe-page 8x11 formula sheet (-sided). No cellphoe or calculator or computer is allowed.

More information

Parametric Density Estimation: Maximum Likelihood Estimation

Parametric Density Estimation: Maximum Likelihood Estimation Parametric Desity stimatio: Maimum Likelihood stimatio C6 Today Itroductio to desity estimatio Maimum Likelihood stimatio Itroducto Bayesia Decisio Theory i previous lectures tells us how to desig a optimal

More information

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

Combining imperfect data, and an introduction to data assimilation Ross Bannister, NCEO, September 2010 Combiig imperfect data, ad a itroductio to data assimilatio Ross Baister, NCEO, September 00 rbaister@readigacuk The probability desity fuctio (PDF prob that x lies betwee x ad x + dx p (x restrictio o

More information

ECON 5350 Class Notes Maximum Likelihood Estimation

ECON 5350 Class Notes Maximum Likelihood Estimation ECON 5350 Class Notes Maximum Likelihood Estimatio 1 Maximum Likelihood Estimatio Example #1. Cosider the radom sample {X 1 = 0.5, X 2 = 2.0, X 3 = 10.0, X 4 = 1.5, X 5 = 7.0} geerated from a expoetial

More information

1 Estimating sensitivities

1 Estimating sensitivities Copyright c 27 by Karl Sigma 1 Estimatig sesitivities Whe estimatig the Greeks, such as the, the geeral problem ivolves a radom variable Y = Y (α) (such as a discouted payoff) that depeds o a parameter

More information

CHAPTER 8 Estimating with Confidence

CHAPTER 8 Estimating with Confidence CHAPTER 8 Estimatig with Cofidece 8.2 Estimatig a Populatio Proportio The Practice of Statistics, 5th Editio Stares, Tabor, Yates, Moore Bedford Freema Worth Publishers Estimatig a Populatio Proportio

More information

Lecture 5 Point Es/mator and Sampling Distribu/on

Lecture 5 Point Es/mator and Sampling Distribu/on Lecture 5 Poit Es/mator ad Samplig Distribu/o Fall 03 Prof. Yao Xie, yao.xie@isye.gatech.edu H. Milto Stewart School of Idustrial Systems & Egieerig Georgia Tech Road map Poit Es/ma/o Cofidece Iterval

More information

Introduction to Probability and Statistics Chapter 7

Introduction to Probability and Statistics Chapter 7 Itroductio to Probability ad Statistics Chapter 7 Ammar M. Sarha, asarha@mathstat.dal.ca Departmet of Mathematics ad Statistics, Dalhousie Uiversity Fall Semester 008 Chapter 7 Statistical Itervals Based

More information

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

Standard Deviations for Normal Sampling Distributions are: For proportions For means _ Sectio 9.2 Cofidece Itervals for Proportios We will lear to use a sample to say somethig about the world at large. This process (statistical iferece) is based o our uderstadig of samplig models, ad will

More information

ii. Interval estimation:

ii. Interval estimation: 1 Types of estimatio: i. Poit estimatio: Example (1) Cosider the sample observatios 17,3,5,1,18,6,16,10 X 8 X i i1 8 17 3 5 118 6 16 10 8 116 8 14.5 14.5 is a poit estimate for usig the estimator X ad

More information

. (The calculated sample mean is symbolized by x.)

. (The calculated sample mean is symbolized by x.) Stat 40, sectio 5.4 The Cetral Limit Theorem otes by Tim Pilachowski If you have t doe it yet, go to the Stat 40 page ad dowload the hadout 5.4 supplemet Cetral Limit Theorem. The homework (both practice

More information

Sampling Distributions and Estimation

Sampling Distributions and Estimation Cotets 40 Samplig Distributios ad Estimatio 40.1 Samplig Distributios 40. Iterval Estimatio for the Variace 13 Learig outcomes You will lear about the distributios which are created whe a populatio is

More information

BASIC STATISTICS ECOE 1323

BASIC STATISTICS ECOE 1323 BASIC STATISTICS ECOE 33 SPRING 007 FINAL EXAM NAME: ID NUMBER: INSTRUCTIONS:. Write your ame ad studet ID.. You have hours 3. This eam must be your ow work etirely. You caot talk to or share iformatio

More information

Inferential Statistics and Probability a Holistic Approach. Inference Process. Inference Process. Chapter 8 Slides. Maurice Geraghty,

Inferential Statistics and Probability a Holistic Approach. Inference Process. Inference Process. Chapter 8 Slides. Maurice Geraghty, Iferetial Statistics ad Probability a Holistic Approach Chapter 8 Poit Estimatio ad Cofidece Itervals This Course Material by Maurice Geraghty is licesed uder a Creative Commos Attributio-ShareAlike 4.0

More information

Maximum Empirical Likelihood Estimation (MELE)

Maximum Empirical Likelihood Estimation (MELE) Maximum Empirical Likelihood Estimatio (MELE Natha Smooha Abstract Estimatio of Stadard Liear Model - Maximum Empirical Likelihood Estimator: Combiatio of the idea of imum likelihood method of momets,

More information

Math 124: Lecture for Week 10 of 17

Math 124: Lecture for Week 10 of 17 What we will do toight 1 Lecture for of 17 David Meredith Departmet of Mathematics Sa Fracisco State Uiversity 2 3 4 April 8, 2008 5 6 II Take the midterm. At the ed aswer the followig questio: To be revealed

More information

Topic-7. Large Sample Estimation

Topic-7. Large Sample Estimation Topic-7 Large Sample Estimatio TYPES OF INFERENCE Ò Estimatio: É Estimatig or predictig the value of the parameter É What is (are) the most likely values of m or p? Ò Hypothesis Testig: É Decidig about

More information

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

Today: Finish Chapter 9 (Sections 9.6 to 9.8 and 9.9 Lesson 3) Today: Fiish Chapter 9 (Sectios 9.6 to 9.8 ad 9.9 Lesso 3) ANNOUNCEMENTS: Quiz #7 begis after class today, eds Moday at 3pm. Quiz #8 will begi ext Friday ad ed at 10am Moday (day of fial). There will be

More information

Chapter 8: Estimation of Mean & Proportion. Introduction

Chapter 8: Estimation of Mean & Proportion. Introduction Chapter 8: Estimatio of Mea & Proportio 8.1 Estimatio, Poit Estimate, ad Iterval Estimate 8.2 Estimatio of a Populatio Mea: σ Kow 8.3 Estimatio of a Populatio Mea: σ Not Kow 8.4 Estimatio of a Populatio

More information

Sampling Distributions & Estimators

Sampling Distributions & Estimators API-209 TF Sessio 2 Teddy Svoroos September 18, 2015 Samplig Distributios & Estimators I. Estimators The Importace of Samplig Radomly Three Properties of Estimators 1. Ubiased 2. Cosistet 3. Efficiet I

More information

Chapter 10 - Lecture 2 The independent two sample t-test and. confidence interval

Chapter 10 - Lecture 2 The independent two sample t-test and. confidence interval Assumptios Idepedet Samples - ukow σ 1, σ - 30 or m 30 - Upooled case Idepedet Samples - ukow σ 1, σ - 30 or m 30 - Pooled case Idepedet samples - Pooled variace - Large samples Chapter 10 - Lecture The

More information

1. Suppose X is a variable that follows the normal distribution with known standard deviation σ = 0.3 but unknown mean µ.

1. Suppose X is a variable that follows the normal distribution with known standard deviation σ = 0.3 but unknown mean µ. Chapter 9 Exercises Suppose X is a variable that follows the ormal distributio with kow stadard deviatio σ = 03 but ukow mea µ (a) Costruct a 95% cofidece iterval for µ if a radom sample of = 6 observatios

More information

CHAPTER 8: CONFIDENCE INTERVAL ESTIMATES for Means and Proportions

CHAPTER 8: CONFIDENCE INTERVAL ESTIMATES for Means and Proportions CHAPTER 8: CONFIDENCE INTERVAL ESTIMATES for Meas ad Proportios Itroductio: I this chapter we wat to fid out the value of a parameter for a populatio. We do t kow the value of this parameter for the etire

More information

Exam 1 Spring 2015 Statistics for Applications 3/5/2015

Exam 1 Spring 2015 Statistics for Applications 3/5/2015 8.443 Exam Sprig 05 Statistics for Applicatios 3/5/05. Log Normal Distributio: A radom variable X follows a Logormal(θ, σ ) distributio if l(x) follows a Normal(θ, σ ) distributio. For the ormal radom

More information

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

18.S096 Problem Set 5 Fall 2013 Volatility Modeling Due Date: 10/29/2013 18.S096 Problem Set 5 Fall 2013 Volatility Modelig Due Date: 10/29/2013 1. Sample Estimators of Diffusio Process Volatility ad Drift Let {X t } be the price of a fiacial security that follows a geometric

More information

ST 305: Exam 2 Fall 2014

ST 305: Exam 2 Fall 2014 ST 305: Exam Fall 014 By hadig i this completed exam, I state that I have either give or received assistace from aother perso durig the exam period. I have used o resources other tha the exam itself ad

More information

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

A random variable is a variable whose value is a numerical outcome of a random phenomenon. The Practice of Statistics, d ed ates, Moore, ad Stares Itroductio We are ofte more iterested i the umber of times a give outcome ca occur tha i the possible outcomes themselves For example, if we toss

More information

The material in this chapter is motivated by Experiment 9.

The material in this chapter is motivated by Experiment 9. Chapter 5 Optimal Auctios The material i this chapter is motivated by Experimet 9. We wish to aalyze the decisio of a seller who sets a reserve price whe auctioig off a item to a group of bidders. We begi

More information

A point estimate is the value of a statistic that estimates the value of a parameter.

A point estimate is the value of a statistic that estimates the value of a parameter. Chapter 9 Estimatig the Value of a Parameter Chapter 9.1 Estimatig a Populatio Proportio Objective A : Poit Estimate A poit estimate is the value of a statistic that estimates the value of a parameter.

More information

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

Exam 2. Instructor: Cynthia Rudin TA: Dimitrios Bisias. October 25, 2011 15.075 Exam 2 Istructor: Cythia Rudi TA: Dimitrios Bisias October 25, 2011 Gradig is based o demostratio of coceptual uderstadig, so you eed to show all of your work. Problem 1 You are i charge of a study

More information

NOTES ON ESTIMATION AND CONFIDENCE INTERVALS. 1. Estimation

NOTES ON ESTIMATION AND CONFIDENCE INTERVALS. 1. Estimation NOTES ON ESTIMATION AND CONFIDENCE INTERVALS MICHAEL N. KATEHAKIS 1. Estimatio Estimatio is a brach of statistics that deals with estimatig the values of parameters of a uderlyig distributio based o observed/empirical

More information

Lecture 4: Probability (continued)

Lecture 4: Probability (continued) Lecture 4: Probability (cotiued) Desity Curves We ve defied probabilities for discrete variables (such as coi tossig). Probabilities for cotiuous or measuremet variables also are evaluated usig relative

More information

Research Article The Probability That a Measurement Falls within a Range of n Standard Deviations from an Estimate of the Mean

Research Article The Probability That a Measurement Falls within a Range of n Standard Deviations from an Estimate of the Mean Iteratioal Scholarly Research Network ISRN Applied Mathematics Volume 0, Article ID 70806, 8 pages doi:0.540/0/70806 Research Article The Probability That a Measuremet Falls withi a Rage of Stadard Deviatios

More information

Basic formula for confidence intervals. Formulas for estimating population variance Normal Uniform Proportion

Basic formula for confidence intervals. Formulas for estimating population variance Normal Uniform Proportion Basic formula for the Chi-square test (Observed - Expected ) Expected Basic formula for cofidece itervals sˆ x ± Z ' Sample size adjustmet for fiite populatio (N * ) (N + - 1) Formulas for estimatig populatio

More information

CHAPTER 8: CONFIDENCE INTERVAL ESTIMATES for Means and Proportions

CHAPTER 8: CONFIDENCE INTERVAL ESTIMATES for Means and Proportions CHAPTER 8: CONFIDENCE INTERVAL ESTIMATES for Meas ad Proportios Itroductio: We wat to kow the value of a parameter for a populatio. We do t kow the value of this parameter for the etire populatio because

More information

Models of Asset Pricing

Models of Asset Pricing APPENDIX 1 TO CHAPTER 4 Models of Asset Pricig I this appedix, we first examie why diversificatio, the holdig of may risky assets i a portfolio, reduces the overall risk a ivestor faces. The we will see

More information

Models of Asset Pricing

Models of Asset Pricing APPENDIX 1 TO CHAPTER4 Models of Asset Pricig I this appedix, we first examie why diversificatio, the holdig of may risky assets i a portfolio, reduces the overall risk a ivestor faces. The we will see

More information

Models of Asset Pricing

Models of Asset Pricing 4 Appedix 1 to Chapter Models of Asset Pricig I this appedix, we first examie why diversificatio, the holdig of may risky assets i a portfolio, reduces the overall risk a ivestor faces. The we will see

More information

These characteristics are expressed in terms of statistical properties which are estimated from the sample data.

These characteristics are expressed in terms of statistical properties which are estimated from the sample data. 0. Key Statistical Measures of Data Four pricipal features which characterize a set of observatios o a radom variable are: (i) the cetral tedecy or the value aroud which all other values are buched, (ii)

More information

Topic 14: Maximum Likelihood Estimation

Topic 14: Maximum Likelihood Estimation Toic 4: November, 009 As before, we begi with a samle X = (X,, X of radom variables chose accordig to oe of a family of robabilities P θ I additio, f(x θ, x = (x,, x will be used to deote the desity fuctio

More information

STAT 135 Solutions to Homework 3: 30 points

STAT 135 Solutions to Homework 3: 30 points STAT 35 Solutios to Homework 3: 30 poits Sprig 205 The objective of this Problem Set is to study the Stei Pheomeo 955. Suppose that θ θ, θ 2,..., θ cosists of ukow parameters, with 3. We wish to estimate

More information

Sampling Distributions and Estimation

Sampling Distributions and Estimation Samplig Distributios ad Estimatio T O P I C # Populatio Proportios, π π the proportio of the populatio havig some characteristic Sample proportio ( p ) provides a estimate of π : x p umber of successes

More information

Confidence Intervals Introduction

Confidence Intervals Introduction Cofidece Itervals Itroductio A poit estimate provides o iformatio about the precisio ad reliability of estimatio. For example, the sample mea X is a poit estimate of the populatio mea μ but because of

More information

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

Online appendices from Counterparty Risk and Credit Value Adjustment a continuing challenge for global financial markets by Jon Gregory Olie appedices from Couterparty Risk ad Credit Value Adjustmet a APPENDIX 8A: Formulas for EE, PFE ad EPE for a ormal distributio Cosider a ormal distributio with mea (expected future value) ad stadard

More information

Unbiased estimators Estimators

Unbiased estimators Estimators 19 Ubiased estimators I Chapter 17 we saw that a dataset ca be modeled as a realizatio of a radom sample from a probability distributio ad that quatities of iterest correspod to features of the model distributio.

More information

Generative Models, Maximum Likelihood, Soft Clustering, and Expectation Maximization

Generative Models, Maximum Likelihood, Soft Clustering, and Expectation Maximization Geerative Models Maximum Likelihood Soft Clusterig ad Expectatio Maximizatio Aris Aagostopoulos We will see why we desig models for data how to lear their parameters ad how by cosiderig a mixture model

More information

Appendix 1 to Chapter 5

Appendix 1 to Chapter 5 Appedix 1 to Chapter 5 Models of Asset Pricig I Chapter 4, we saw that the retur o a asset (such as a bod) measures how much we gai from holdig that asset. Whe we make a decisio to buy a asset, we are

More information

= α e ; x 0. Such a random variable is said to have an exponential distribution, with parameter α. [Here, view X as time-to-failure.

= α e ; x 0. Such a random variable is said to have an exponential distribution, with parameter α. [Here, view X as time-to-failure. 1 Homewor 1 AERE 573 Fall 018 DUE 8/9 (W) Name ***NOTE: A wor MUST be placed directly beeath the associated part of a give problem.*** PROBEM 1. (5pts) [Boo 3 rd ed. 1.1 / 4 th ed. 1.13] et ~Uiform[0,].

More information

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

FINM6900 Finance Theory How Is Asymmetric Information Reflected in Asset Prices? FINM6900 Fiace Theory How Is Asymmetric Iformatio Reflected i Asset Prices? February 3, 2012 Referece S. Grossma, O the Efficiecy of Competitive Stock Markets where Traders Have Diverse iformatio, Joural

More information

of Asset Pricing R e = expected return

of Asset Pricing R e = expected return Appedix 1 to Chapter 5 Models of Asset Pricig EXPECTED RETURN I Chapter 4, we saw that the retur o a asset (such as a bod) measures how much we gai from holdig that asset. Whe we make a decisio to buy

More information

Non-Inferiority Logrank Tests

Non-Inferiority Logrank Tests Chapter 706 No-Iferiority Lograk Tests Itroductio This module computes the sample size ad power for o-iferiority tests uder the assumptio of proportioal hazards. Accrual time ad follow-up time are icluded

More information

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

r i = a i + b i f b i = Cov[r i, f] The only parameters to be estimated for this model are a i 's, b i 's, σe 2 i The iformatio required by the mea-variace approach is substatial whe the umber of assets is large; there are mea values, variaces, ad )/2 covariaces - a total of 2 + )/2 parameters. Sigle-factor model:

More information

B = A x z

B = A x z 114 Block 3 Erdeky == Begi 6.3 ============================================================== 1 / 8 / 2008 1 Correspodig Areas uder a ormal curve ad the stadard ormal curve are equal. Below: Area B = Area

More information

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

Summary. Recap. Last Lecture. .1 If you know MLE of θ, can you also know MLE of τ(θ) for any function τ? Last Lecture Biostatistics 60 - Statistical Iferece Lecture Cramer-Rao Theorem Hyu Mi Kag February 9th, 03 If you kow MLE of, ca you also kow MLE of τ() for ay fuctio τ? What are plausible ways to compare

More information

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

INTERVAL GAMES. and player 2 selects 1, then player 2 would give player 1 a payoff of, 1) = 0. INTERVAL GAMES ANTHONY MENDES Let I ad I 2 be itervals of real umbers. A iterval game is played i this way: player secretly selects x I ad player 2 secretly ad idepedetly selects y I 2. After x ad y are

More information

Subject CT1 Financial Mathematics Core Technical Syllabus

Subject CT1 Financial Mathematics Core Technical Syllabus Subject CT1 Fiacial Mathematics Core Techical Syllabus for the 2018 exams 1 Jue 2017 Subject CT1 Fiacial Mathematics Core Techical Aim The aim of the Fiacial Mathematics subject is to provide a groudig

More information

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

of Asset Pricing APPENDIX 1 TO CHAPTER EXPECTED RETURN APPLICATION Expected Return APPENDIX 1 TO CHAPTER 5 Models of Asset Pricig I Chapter 4, we saw that the retur o a asset (such as a bod) measures how much we gai from holdig that asset. Whe we make a decisio to buy a asset, we are

More information

Control Charts for Mean under Shrinkage Technique

Control Charts for Mean under Shrinkage Technique Helderma Verlag Ecoomic Quality Cotrol ISSN 0940-5151 Vol 24 (2009), No. 2, 255 261 Cotrol Charts for Mea uder Shrikage Techique J. R. Sigh ad Mujahida Sayyed Abstract: I this paper a attempt is made to

More information

AY Term 2 Mock Examination

AY Term 2 Mock Examination AY 206-7 Term 2 Mock Examiatio Date / Start Time Course Group Istructor 24 March 207 / 2 PM to 3:00 PM QF302 Ivestmet ad Fiacial Data Aalysis G Christopher Tig INSTRUCTIONS TO STUDENTS. This mock examiatio

More information

Overlapping Generations

Overlapping Generations Eco. 53a all 996 C. Sims. troductio Overlappig Geeratios We wat to study how asset markets allow idividuals, motivated by the eed to provide icome for their retiremet years, to fiace capital accumulatio

More information

Problem Set 1a - Oligopoly

Problem Set 1a - Oligopoly Advaced Idustrial Ecoomics Sprig 2014 Joha Steek 6 may 2014 Problem Set 1a - Oligopoly 1 Table of Cotets 2 Price Competitio... 3 2.1 Courot Oligopoly with Homogeous Goods ad Differet Costs... 3 2.2 Bertrad

More information

43. A 000 par value 5-year bod with 8.0% semiaual coupos was bought to yield 7.5% covertible semiaually. Determie the amout of premium amortized i the 6 th coupo paymet. (A).00 (B).08 (C).5 (D).5 (E).34

More information

Department of Mathematics, S.R.K.R. Engineering College, Bhimavaram, A.P., India 2

Department of Mathematics, S.R.K.R. Engineering College, Bhimavaram, A.P., India 2 Skewess Corrected Cotrol charts for two Iverted Models R. Subba Rao* 1, Pushpa Latha Mamidi 2, M.S. Ravi Kumar 3 1 Departmet of Mathematics, S.R.K.R. Egieerig College, Bhimavaram, A.P., Idia 2 Departmet

More information

Twitter: @Owe134866 www.mathsfreeresourcelibrary.com Prior Kowledge Check 1) State whether each variable is qualitative or quatitative: a) Car colour Qualitative b) Miles travelled by a cyclist c) Favourite

More information

0.1 Valuation Formula:

0.1 Valuation Formula: 0. Valuatio Formula: 0.. Case of Geeral Trees: q = er S S S 3 S q = er S S 4 S 5 S 4 q 3 = er S 3 S 6 S 7 S 6 Therefore, f (3) = e r [q 3 f (7) + ( q 3 ) f (6)] f () = e r [q f (5) + ( q ) f (4)] = f ()

More information

1 Estimating the uncertainty attached to a sample mean: s 2 vs.

1 Estimating the uncertainty attached to a sample mean: s 2 vs. Political Sciece 100a/200a Fall 2001 Cofidece itervals ad hypothesis testig, Part I 1 1 Estimatig the ucertaity attached to a sample mea: s 2 vs. σ 2 Recall the problem of descriptive iferece: We wat to

More information

4.5 Generalized likelihood ratio test

4.5 Generalized likelihood ratio test 4.5 Geeralized likelihood ratio test A assumptio that is used i the Athlete Biological Passport is that haemoglobi varies equally i all athletes. We wish to test this assumptio o a sample of k athletes.

More information

BIOSTATS 540 Fall Estimation Page 1 of 72. Unit 6. Estimation. Use at least twelve observations in constructing a confidence interval

BIOSTATS 540 Fall Estimation Page 1 of 72. Unit 6. Estimation. Use at least twelve observations in constructing a confidence interval BIOSTATS 540 Fall 015 6. Estimatio Page 1 of 7 Uit 6. Estimatio Use at least twelve observatios i costructig a cofidece iterval - Gerald va Belle What is the mea of the blood pressures of all the studets

More information

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

Chpt 5. Discrete Probability Distributions. 5-3 Mean, Variance, Standard Deviation, and Expectation Chpt 5 Discrete Probability Distributios 5-3 Mea, Variace, Stadard Deviatio, ad Expectatio 1/23 Homework p252 Applyig the Cocepts Exercises p253 1-19 2/23 Objective Fid the mea, variace, stadard deviatio,

More information

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

Confidence Intervals. CI for a population mean (σ is known and n > 30 or the variable is normally distributed in the. Cofidece Itervals A cofidece iterval is a iterval whose purpose is to estimate a parameter (a umber that could, i theory, be calculated from the populatio, if measuremets were available for the whole populatio).

More information

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

STRAND: FINANCE. Unit 3 Loans and Mortgages TEXT. Contents. Section. 3.1 Annual Percentage Rate (APR) 3.2 APR for Repayment of Loans CMM Subject Support Strad: FINANCE Uit 3 Loas ad Mortgages: Text m e p STRAND: FINANCE Uit 3 Loas ad Mortgages TEXT Cotets Sectio 3.1 Aual Percetage Rate (APR) 3.2 APR for Repaymet of Loas 3.3 Credit Purchases

More information

1 Random Variables and Key Statistics

1 Random Variables and Key Statistics Review of Statistics 1 Radom Variables ad Key Statistics Radom Variable: A radom variable is a variable that takes o differet umerical values from a sample space determied by chace (probability distributio,

More information

Lecture 5: Sampling Distribution

Lecture 5: Sampling Distribution Lecture 5: Samplig Distributio Readigs: Sectios 5.5, 5.6 Itroductio Parameter: describes populatio Statistic: describes the sample; samplig variability Samplig distributio of a statistic: A probability

More information

Calculation of the Annual Equivalent Rate (AER)

Calculation of the Annual Equivalent Rate (AER) Appedix to Code of Coduct for the Advertisig of Iterest Bearig Accouts. (31/1/0) Calculatio of the Aual Equivalet Rate (AER) a) The most geeral case of the calculatio is the rate of iterest which, if applied

More information

Simulation Efficiency and an Introduction to Variance Reduction Methods

Simulation Efficiency and an Introduction to Variance Reduction Methods Mote Carlo Simulatio: IEOR E4703 Columbia Uiversity c 2017 by Marti Haugh Simulatio Efficiecy ad a Itroductio to Variace Reductio Methods I these otes we discuss the efficiecy of a Mote-Carlo estimator.

More information

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

A New Constructive Proof of Graham's Theorem and More New Classes of Functionally Complete Functions A New Costructive Proof of Graham's Theorem ad More New Classes of Fuctioally Complete Fuctios Azhou Yag, Ph.D. Zhu-qi Lu, Ph.D. Abstract A -valued two-variable truth fuctio is called fuctioally complete,

More information

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

Rafa l Kulik and Marc Raimondo. University of Ottawa and University of Sydney. Supplementary material Statistica Siica 009: Supplemet 1 L p -WAVELET REGRESSION WITH CORRELATED ERRORS AND INVERSE PROBLEMS Rafa l Kulik ad Marc Raimodo Uiversity of Ottawa ad Uiversity of Sydey Supplemetary material This ote

More information

An Improved Estimator of Population Variance using known Coefficient of Variation

An Improved Estimator of Population Variance using known Coefficient of Variation J. Stat. Appl. Pro. Lett. 4, No. 1, 11-16 (017) 11 Joural of Statistics Applicatios & Probability Letters A Iteratioal Joural http://dx.doi.org/10.18576/jsapl/04010 A Improved Estimator of Populatio Variace

More information

An Empirical Study of the Behaviour of the Sample Kurtosis in Samples from Symmetric Stable Distributions

An Empirical Study of the Behaviour of the Sample Kurtosis in Samples from Symmetric Stable Distributions A Empirical Study of the Behaviour of the Sample Kurtosis i Samples from Symmetric Stable Distributios J. Marti va Zyl Departmet of Actuarial Sciece ad Mathematical Statistics, Uiversity of the Free State,

More information

Statistics for Business and Economics

Statistics for Business and Economics Statistics for Busiess ad Ecoomics Chapter 8 Estimatio: Additioal Topics Copright 010 Pearso Educatio, Ic. Publishig as Pretice Hall Ch. 8-1 8. Differece Betwee Two Meas: Idepedet Samples Populatio meas,

More information

Confidence Intervals based on Absolute Deviation for Population Mean of a Positively Skewed Distribution

Confidence Intervals based on Absolute Deviation for Population Mean of a Positively Skewed Distribution Iteratioal Joural of Computatioal ad Theoretical Statistics ISSN (220-59) It. J. Comp. Theo. Stat. 5, No. (May-208) http://dx.doi.org/0.2785/ijcts/0500 Cofidece Itervals based o Absolute Deviatio for Populatio

More information

A New Approach to Obtain an Optimal Solution for the Assignment Problem

A New Approach to Obtain an Optimal Solution for the Assignment Problem Iteratioal Joural of Sciece ad Research (IJSR) ISSN (Olie): 231-7064 Idex Copericus Value (2013): 6.14 Impact Factor (2015): 6.31 A New Approach to Obtai a Optimal Solutio for the Assigmet Problem A. Seethalakshmy

More information

Chapter 8 Interval Estimation. Estimation Concepts. General Form of a Confidence Interval

Chapter 8 Interval Estimation. Estimation Concepts. General Form of a Confidence Interval Chapter 8 Iterval Estimatio Estimatio Cocepts Usually ca't take a cesus, so we must make decisios based o sample data It imperative that we take the risk of samplig error ito accout whe we iterpret sample

More information

EXERCISE - BINOMIAL THEOREM

EXERCISE - BINOMIAL THEOREM BINOMIAL THOEREM / EXERCISE - BINOMIAL THEOREM LEVEL I SUBJECTIVE QUESTIONS. Expad the followig expressios ad fid the umber of term i the expasio of the expressios. (a) (x + y) 99 (b) ( + a) 9 + ( a) 9

More information

Notes on Expected Revenue from Auctions

Notes on Expected Revenue from Auctions Notes o Epected Reveue from Auctios Professor Bergstrom These otes spell out some of the mathematical details about first ad secod price sealed bid auctios that were discussed i Thursday s lecture You

More information

We analyze the computational problem of estimating financial risk in a nested simulation. In this approach,

We analyze the computational problem of estimating financial risk in a nested simulation. In this approach, MANAGEMENT SCIENCE Vol. 57, No. 6, Jue 2011, pp. 1172 1194 iss 0025-1909 eiss 1526-5501 11 5706 1172 doi 10.1287/msc.1110.1330 2011 INFORMS Efficiet Risk Estimatio via Nested Sequetial Simulatio Mark Broadie

More information

CAPITAL PROJECT SCREENING AND SELECTION

CAPITAL PROJECT SCREENING AND SELECTION CAPITAL PROJECT SCREEIG AD SELECTIO Before studyig the three measures of ivestmet attractiveess, we will review a simple method that is commoly used to scree capital ivestmets. Oe of the primary cocers

More information

Solutions to Problem Sheet 1

Solutions to Problem Sheet 1 Solutios to Problem Sheet ) Use Theorem.4 to prove that p log for all real x 3. This is a versio of Theorem.4 with the iteger N replaced by the real x. Hit Give x 3 let N = [x], the largest iteger x. The,

More information

Monetary Economics: Problem Set #5 Solutions

Monetary Economics: Problem Set #5 Solutions Moetary Ecoomics oblem Set #5 Moetary Ecoomics: oblem Set #5 Solutios This problem set is marked out of 1 poits. The weight give to each part is idicated below. Please cotact me asap if you have ay questios.

More information

Introduction to Econometrics (3 rd Updated Edition) Solutions to Odd- Numbered End- of- Chapter Exercises: Chapter 2

Introduction to Econometrics (3 rd Updated Edition) Solutions to Odd- Numbered End- of- Chapter Exercises: Chapter 2 Itroductio to Ecoometrics (3 rd Updated Editio) by James H. Stock ad Mark W. Watso Solutios to Odd- Numbered Ed- of- Chapter Exercises: Chapter 2 (This versio August 7, 204) Stock/Watso - Itroductio to

More information

Kernel Density Estimation. Let X be a random variable with continuous distribution F (x) and density f(x) = d

Kernel Density Estimation. Let X be a random variable with continuous distribution F (x) and density f(x) = d Kerel Desity Estimatio Let X be a radom variable wit cotiuous distributio F (x) ad desity f(x) = d dx F (x). Te goal is to estimate f(x). Wile F (x) ca be estimated by te EDF ˆF (x), we caot set ˆf(x)

More information