MAY 2005 SOA EXAM C/CAS 4 SOLUTIONS

Size: px
Start display at page:

Download "MAY 2005 SOA EXAM C/CAS 4 SOLUTIONS"

Transcription

1 MAY 2005 SOA EXAM C/CAS 4 SOLUTIONS Prepared by Sam Broverman, to appear in ACTEX Exam C/4 Study Guide 2brove@rogerscom sam@utstattorontoedu 1 The distribution function is JÐBÑ ' B 0Ð>Ñ> ' B %!! > Ð>Ñ& Ð>Ñ% The Kolmogorov-Smirnov statistic is Q+BÖlJ ÐB Ñ J ÐB Ñl ß lj ÐB Ñ J ÐB Ñl where the maximum is taken over all data points JÐB ÑJÐB Ñ J 8 is the empirical distribution function B3 J ÐB3Ñ J8ÐB3 Ñß J8ÐB3Ñ lj ÐB3Ñ J8ÐB3 Ñl lj ÐB3Ñ J8ÐB3Ñl Þ Þ&((! ß Þ Þ&(( Þ(( Þ( Þ! Þ ß Þ% Þ'! Þ%! Þ* Þ* Þ% ß Þ' Þ& Þ Þ Þ*%' Þ' ß Þ Þ%' Þ%' Þ Þ*'% Þ ß Þ! Þ'% Þ!&( The K-S statistic is 6803 With a sample of size 5, the critical values are Level of Significance Critical Value The K-S statistic is Þ''* so L! is rejected at the 025 significance level, but the statistic is Þ(*!, so L is not rejected at the 01 significance level Answer: D! 2 We are using the full credibility standard based on the number of claims needed for full credibility of aggregate losses of a compound distribution W This standard is Þ*' Ð Þ Ñ Z+<ÒWÓ ÐIÒWÓÑ IÒRÓ, where W is aggregate losses and R is the frequency distribution (negative binomial in this case The factor 196 comes from the 95% probability requirement (196 is the 975-percentile of the standard normal With a severity random variable ], the mean of the compound distribution W is IÒWÓ IÒRÓ IÒ] Ó and the variance is Z+<ÒWÓIÒR ] Z+<Ò]ÓZ+<ÒRÓ ÐIÒ]ÓÑ For this problem, IÒRÓ< < and Z+<ÒRÓ< Ð Ñ<, IÒ] Ó ÐÑÐÞ%Ñ Ð!ÑÐÞ%Ñ Ð!!ÑÐÞÑ %Þ%, IÒ] Ó ÐÑ ÐÞ%Ñ Ð!Ñ ÐÞ%Ñ Ð!!Ñ ÐÞÑ!%!Þ%, and Z +<Ò] Ó!%!Þ% Ð%Þ%Ñ %%&Þ!% Then, IÒWÓ Ð<ÑÐ%Þ%Ñ (Þ< and Z +<ÒWÓ Ð<ÑÐ%%&Þ!%Ñ Ð<ÑÐ%Þ%Ñ ß %(*Þ%%< ß%(*Þ%%< The standard for full credibility becomes Ð%Þ'Ñ Ð(Þ<Ñ Ð<Ñ %'*Þ The expected number of claims needed is at least 2470 Answer: E 2brove@rogerscom sam@utstattorontoedu

2 3 There is no censoring and there is exactly one death at each time of death, and therefore this is a complete data set If there are 8 individuals alive at time!, the number at risk at the first death point is 8 The number at risk at the second death point is 8, etc The product-limit estimate 85 of survival to the 5-th time of death is Ð 8 ÑÐ 8 ÑâÐ 85 Ñ 8, so that 8* WÐ>Ñ 8 * 8 If we knew the value of 8, we could find WÐ>Ñ 8 * The Nelson-Aalen estimate of LÐ> Ñ is LÐ> s = = * Ñ < < 8 8! From inspection it can be seen that 8! (alternatively, 8 8 is approximately equal to *, and if we solve 8 8!, we get 8 *Þ**, but 8 must be an integer!* Then, W8 Ð>ÑW *! Ð>Ñ *! Þ&& Answer: A 4 The empirical distribution is a 2-point random variable with :ÐÑ and :Ð%Ñ The mean of the empirical distribution is ÐÑÐ Ñ Ð%ÑÐ Ñ Þ The third central moment of the empirical distribution is ÐÑ Ð ÑÐ%Ñ Ð Ñ The estimator is s 1Ð\ ß\ ß\ Ñ D Ð\ 3 \Ñ for a sample \ ß\ ß\ Þ There are ( possible bootstrap samples (of size 3 from the original sample For each sample we calculate the estimate s based on the values for that sample These 27 samples can be described as follows: - all 1's, 8 samples, \ß s!ß Ðs Ñ Ð!Ñ %à - two 1's and one %, 12 samples, \ßs ÒÐÑ ÐÑ Ð%Ñ Óß and Ðs Ñ ÐÑ! à - one 1 and two %'s, 6 samples, \ßs ÒÐÑ Ð%Ñ Ð%Ñ Ó ß and Ðs Ñ ÐÑ ' à - all %'s, 1 sample, \ % ß s! ß Ðs Ñ Ð!Ñ % The bootstrap estimate to the MSE of the estimator is ' Ð ( ÑÐ%Ñ Ð ( ÑÐ!Ñ Ð ( ÑÐ'Ñ Ð ( ÑÐ%Ñ ( %Þ* An example of the count of the number of samples given above is the number of samples with one 1 and two 4's: Ðß4ß4Ñß Ðß4ß4Ñß Ð4ßß4Ñß Ð4ßß4Ñß Ð4ß4ßÑß Ð4ß4 ßÑ The reason that there are two samples of the form Ðß4ß4 Ñ is that there are two sample points equal to 1 If we label them 1a and 1b, then Ð+ß 4ß 4 Ñ and Ð,ß 4ß 4 should be regarded as separate samples The same applies to the other samples Answer: E 2brove@rogerscom sam@utstattorontoedu

3 5 The sample is of size 9 For the :: - plot, we assign smoothed percentiles to the sample points &! &! & **!! % & ' ( * Smoothed!!!!!!!!! Percentile J8ÐBÑ We then find JÐB3Ñfor the fitted cdf J and the :-: plot has the smoothed percentile as the horizontal coordinate and the fitted cdf value JÐB Ñ 3 as the vertical coordinate We can determine which is the correct fitted model by elimination For Model B, J ÐÑ Î, but in the plot, we have J ÐÑ! (this is the vertical coordinate of the first plotted point This eliminates B B For Model C, the uniform on Òß!!Ó, J ÐBÑ ** ß so J Ð!!Ñ, but in the plot we have J Ð!!Ñ Þ( (approx(this is the vertical coordinate of the last plotted point This eliminates C BÎ! For Model D, the cdf is J ÐBÑ /, so that J ÐÑ Þ!*&, but in the plot, we have JÐÑ! This eliminates D B%! For Model E, J ÐBÑ FÐ %! Ñ, so that J Ð!Ñ FÐ Þ&Ñ Þ&, but in the plot we have J Ð!Ñ Þ& This eliminates Model E Þ& Model A can also be seen to fit by considering JÐB Ñ B for each B : &! &! & **!! % & ' ( * Smoothed!!!!!!!!! Percentile J8ÐBÑ B 3Þ&! Þ' Þ% Þ%* Þ&( Þ' Þ' Þ' Þ' These are the vertical coordinates in the plot Answer: A 6 With 8& observations, the Buhlmann credibility factor is ^ & + We are told that \l is Poisson with mean Therefore, the hypothetical mean Ð Ñ IÒ\l@ and the process variance Ñ Z+<Ò\l@ From the form of the cdf we see has a Pareto distribution with Þ' and (not the of the prior - same letter, different variable Therefore + Z +<Ò@ Ð ÑÓ Z +<Ò@ Ó ÐÞ'ÑÐÞ'Ñ Ð Þ' Ñ Þ'*(, IÒZ +<Ò\l@ ÓÓ IÒ@ Ó Þ' Þ'& Þ & Finally, ^ Þ* Answer: E & Þ'& Þ'*( 2brove@rogerscom sam@utstattorontoedu

4 7 Using the notation associated with the Kaplan-Meier approximation for large data sets, we have -!!! ß - &!! ß -!!! ß - &!!! ß -%!ß!!!! %!! (there are 400 observations with deductible of 300, && deducible of 525,! % (observations with (there are no observations with deducible higher than 5000 B &!, B & ß B!! ß B!! ß?! ß? %! Þ! 4 4 With and!, the number at risk at - is < ÐB?Ñ, and we use the estimator JÐ- s B! B4 4 Ñ Š < â Š < Then with! 4 - &!!!, we have <!! %!! ß ! 3! < Ð Ñ ÐB? Ñ Ð%!! &&Ñ Ð&!!Ñ (& ß!!! < Ð ÑÐB B?? ÑÐ%!!&&!ÑÐ&!&!!Ñ(&! Then Js B! B B Ð&!!!Ñ Š < Š < Š <! &! &!! Š %!! Š (& Š (&! Þ&& Answer: E!!! 8 The two knots are ÐB! ß C! Ñ Ð!ß!Ñ and ÐB ß C Ñ Ðß!Ñ The cubic spline on the two knots can be written in the form 0ÐBÑ+!,B-B!! B!,!ŸBŸÞ The spline must satisfy the relationships 0ÐB! Ñ C!, from which we get +!!, and 0ÐB Ñ C, from which we get,! %-!!! (A w w!!!!!! Also, 0 Ð!Ñ,, and 0 ÐÑ, - ÐÑ ÐÑ %- (B Equations A and B become %-!! % and %-!! % Solving this system results in -! and!!, along with,! and +!! The spline function is 0ÐBÑ BB!! The squared norm measure is ' ww ' Ò0 ÐBÑÓ B B Answer: D 9 If \ has an exponential distribution with mean, and \ and \ forms a random sample of size two, then \ \ has a gamma distribution with parameters and the same as in the exponential distribution If ] sample mean of \ and \, then >/ >Î T Ò]!Ó T Ò\ \!Ó '! > This integral can be found by integration by parts once is known Since the original random variable is exponential, and we have a sample of size two, the maximum likelihood estimate of is the sample mean, which is 6 The maximum likelihood estimate of the probability above is >Î' >Î' >!Î' ' >/ >/ >Î'!/!Î'! ' > ' / ¹! Ò >! ' / Ó Þ& +> +> We have used the integration by parts relationship ' +> >/ / >/ > + + ; in this case + ' Answer: D 2brove@rogerscom sam@utstattorontoedu

5 ! >/ >Î 10 From 9 above we have J] Ð!Ñ T Ò] Ÿ!Ó T Ò\ \ Ÿ!Ó '! > Using the mle s! >Îs, the estimated probability is ' >/ > 1Ðs! Ñ s Using the delta method, the estimated variance of 1ÐsÑ is Z +<Ð1Ðs w ÑÑ Ò1 ÐIÐs ÑÑÓ Z +<Òs Ó Since s is the mle of the mean of the exponential distribution, we know that s B, which is the sample mean of the first two observed values Therefore, Z+<ÒsÓZ+<ÒBÓ Z+<Ò\Ó (the variance of the exponential distribution is the square of the mean Using the antiderivative formula mentioned in the solution to 9 above, we have >! >Îs >Îs!Îs! ' >/ >/ >Îs!/!Îs! > s s / ¹ s /! >!!Îs!/!Îs / 1Ðs s Ñ! s!îs!!îs / Ð Ñ!/ w Then, 1Ðs s!îs! Ñ / Ð Ñ s s Using the estimate value s ', we get!î'!!î'!ð'ñ/ Ð Ñ!/ w '!Î' 1 Ð'Ñ! ' / Ð ' Ñ Þ!'' Then Z+<Ð1Ðs w ÑÑis approximately equal to Ò1Ðs ÑÓ Z+<Òs w ÓÒ1Ðs ÑÓ, which is w ' estimated to be Ò1 Ð'ÑÓ ÐÞ!''Ñ ÐÑ Þ!(* Answer: A 11 Credibility is being applied to aggregate losses W, which has a compound distribution The frequency R depends on the parameter -, and the severity ] depends on the parameter, so W depends on both parameters In general, for a compound distribution, the mean is IÒWÓIÒRÓ IÒ]Ó and the variance is Z+<ÒWÓIÒRÓ Z+<Ò]ÓZ+<ÒRÓ ÐIÒ]ÓÑ The hypothetical mean in this example is IÒWl- ß Ó IÒRl- Ó IÒ] l Ó - and the process variance is Z +<ÒWl- ß Ó IÒRl- Ó Z +<Ò] l Ó Z +<ÒRl- Ó ÐIÒ] l ÓÑ The variance of the hypothetical mean is +Z+<ÒIÒWl ß-ÓÓZ+<Ò- ÓIÒÐ- Ñ ÓÐIÒ- ÓÑ Since - and are independent, we have IÒ- Ó IÒ- Ó IÒ Ó, and IÒÐ- Ñ Ó IÒ- Ó IÒ- Ó IÒ Ó ÐÑÐÑ % (since - has an exponential distribution with mean 1, the second moment of - is Ðmean Ñ, and since has a Poisson distribution with mean 1, Z +<Ò Ó IÒ Ó ÐIÒ ÓÑ IÒ Ó ÐÑ, so that IÒ Ó Therefore, +% The expected process variance IÒZ +<ÒWl- ß ÓÓ IÒ - Ó IÒ- Ó IÒ Ó ÐÑÐÑ % Then 5 + Answer: B 2brove@rogerscom sam@utstattorontoedu

6 12 The number of deaths R, from 100 people has a binomial distribution with 7!! and ;!Þ! The inversion method requires knowing the distribution function The probability!! 5!!5 function for the binomial is T ÒR 5Ó Š 5 ÐÞ!Ñ ÐÞ*(Ñ 5! ÞÞÞ T ÒR 5Ó Þ!%(&& Þ%(!( Þ&& J Ð5Ñ Þ!%(&& Þ*%' Þ%*(( From? Þ!, the simulated value of R is 8, since Þ*%' Ÿ Þ Þ%*(( Þ From? Þ!, the simulated value of R is 8!, since Þ! Þ!%(&& Þ From? Þ!*, the simulated value of R is 8, since Þ!%(&& Ÿ Þ!* Þ*%' Þ! The average of the simulated values is Answer: B 13 If ] is a mixture of \ and \ with mixing weights + and +, we can define the Öß, with T Ò@ Ó + ß T Ò@ Ó + Then IÒ]ÓIÒIÒ]l@ ÓÓ IÒ]l@ Ó TÒ@ ÓIÒ]l@ Ó TÒ@ Ó IÒ\ Ó +IÒ\ Ó Ð+Ñ (this is the usual way the mean of a finite mixture is formulated Z+<Ò]ÓIÒZ+<Ò]l@ ÓÓZ+<ÒIÒ]l@ ÓÓ, where IÒZ+<Ò]l@ ÓÓZ+<Ò\ Ó +Z+<Ò\ Ó Ð+Ñ and Z +<ÒIÒ] l@ ÓÓ ÐIÒ\ Ó IÒ\ ÓÑ +Ð +Ñ The last equality follows from the fact that if ^ is a two-point random variable? prob : ^š, then Z+<Ò^ÓÐ?@Ñ :Ð:Ñ; IÒ]l@ prob : prob a two-point random variable IÒ] l Ó š Ó + IÒ\ Ó prob + Therefore Z +<Ò] Ó Z +<Ò\ Ó + Z +<Ò\ Ó Ð +Ñ ÐIÒ\ Ó IÒ\ ÓÑ +Ð +Ñ B If \ß\ are Poisson random variables then IÒ\ÓZ+<Ò\Ó and IÒ\ÓZ+<Ò\Ó, so that IÒ]ÓIÒ\ Ó +IÒ\ Ó Ð+ÑZ+<Ò\ Ó +Z+<Ò\ Ó Ð+Ñ It follows that Z +<Ò] Ó Z +<Ò\ Ó + Z +<Ò\ Ó Ð +Ñ ÐIÒ\ Ó IÒ\ ÓÑ +Ð +Ñ IÒ] Ó C For a negative binomial random variable with parameters < and, the mean is < and the variance is < Ð Ñ, so the variance is larger than the mean If \ and \ have negative binomial distributions, the IÒ\ Ó Z +<Ò\ Ó and IÒ\ Ó Z +<Ò\ Ó 2brove@rogerscom sam@utstattorontoedu

7 Therefore, IÒ]ÓIÒ\ Ó +IÒ\ Ó Ð+ÑZ+<Ò\ Ó +Z+<Ò\ Ó Ð+Ñ, and Z+<Ò\ Ó +Z+<Ò\ Ó Ð+Ñ Z +<Ò\ Ó + Z +<Ò\ Ó Ð +Ñ ÐIÒ\ Ó IÒ\ ÓÑ +Ð +Ñ Z +<Ò] Ó ; therefore IÒ] Ó Z +<Ò] Ó A For a binomial random variable with parameters 7 and ;, the mean is 7; and the variance is 7;Ð ;Ñ, which is smaller than the mean Therefore IÒ]ÓIÒ\ Ó +IÒ\ Ó Ð+ÑZ+<Ò\ Ó +Z+<Ò\ Ó Ð+Ñ, and it is possible that when we add ÐIÒ\ Ó IÒ\ ÓÑ +Ð +Ñ to the right side, we get approximate equality So it is possible that IÒ] Ó Z +<Ò] Ó Answer: A for a mixture of binomials 14 We wish to find IÒ\ l\!ó This can be formulated as '! IÒ\ l-ó 1Ð-l\!Ñ -, where 1Ð-l\!Ñ is the posterior density of - given 0Ð!ß-Ñ \! The posterior density is 1- Ð l\!ñ 0 Ð!Ñ, where the numerator is the joint density of \ and -, and the denominator is the marginal probability that \! / The joint density is 0Ð!ß -Ñ 0Ð!l-Ñ 1Ð-Ñ -! - -Î' -Î!x ÒÐ!Þ%Ñ ' / Ð!Þ'Ñ / Ó Þ%! ( -Î' Þ'! -Î This can be written as 0Ð!ß -Ñ '!x - /!x - / The marginal probability 0\ Ð!Ñ is 0\ Ð!Ñ ' 0Ð!ß Ñ ' Þ%!! Ò! / - Î' Þ' ! / - '!x!x Î Ó - Þ%!x Þ'!x Þ% ' Þ' '!x Ð(Î'Ñ!x ÐÎÑ ' Ð ( Ñ Ð Ñ We have used the relationship '! >/ 5 +> > 5x for integer 5! The posterior density is Þ%! ( -Î' Þ'! - Î Þ% ' Þ' 1- Ð l\!ñ Ò '!x - /!x - / Ó Ò ' Ð ( Ñ Ð Ñ Ó The Bayesian premium IÒ\ l\!ó is '! - 1Ð-l\!Ñ - Using the integral relationship mentioned above, we have ' Þ%! ( -Î' Þ'! - Î! - Ò '!x - /!x - / Ó - Þ% x Þ' x ÐÞ%ÑÐÑ ' ÐÞ'ÑÐÑ '!x Ð(Î'Ñ!x ÐÎÑ ' Ð ( Ñ Ð Ñ ÐÞ%ÑÐÑ ' ÐÞ'ÑÐÑ ' Ð ( Ñ Ð Ñ Finally, IÒ\ l\!ó Þ% ' Þ' *Þ Ð Ñ Ð Ñ ' ( + 5 \ 2brove@rogerscom sam@utstattorontoedu

8 14 continued Another approach to solving this problem follows from a careful look at the algebraic form of the Þ%! ( -Î' Þ'! -Î joint density 0Ð!ß -Ñ '!x - /!x - / We know that the posterior density is proportional to this (as far as the parameter - is concerned, so the posterior must be a mixture of two gamma distributions The first gamma in the posterior has ß ' (, and the second gamma has ß Therefore, the posterior density will be of the form - +! ( - Î' - Ð+Ñ! - / / Î ' Ð Since the posterior density is proportional to the joint ( Ñ!x Ð Ñ!x density, the ratio of the factors Ò Þ' Ó Ò Þ% Ó!x Î '!x % in the joint density, must be the same as + + ' Ð+Ñ the ratio of the factors Ò Ó Ò Ó Ð Ñ Î '!x Ð Ñ!x ( + ( ' Therefore, ( Ð + Ñ %, from which we get + Þ( ß + Þ'* Then IÒ\ l\!ó is '! - 1Ð-l\!Ñ - is the mean of the posterior distribution, which is the mixture of the two means + Ð +Ñ *Þ Answer: D 15 The interval is LÐ%Þ&Ñ s Þ*' ÉZ s+<òlð%þ&ñó s There is no indication of any censoring or truncation (monitoring is from starting date of policy to time of first claim, so the number at risk is reduced only by the numbers of claims at the various claim time points LÐ%Þ&Ñ s! * ( Þ((%' (the most recent claim time before 45 is time 4 The Aalen estimate of the variance of LÐ%Þ&Ñ s is Z s+<òlð%þ&ñó s! * ( Þ!*% The interval is Þ((%' Þ*' ÈÞ!*% ÐÞ* ß Þ'Ñ Answer: A 16 For any estimator s we have the relationship s QWIs Ð Ñ Z+<Ò ÓÒ,3+= s Ð ÑÓ From (iv it follows that Z+<Òs ÓÒ,3+= s Ð ÑÓ 5 Since s 5 \, it follows that,3+= Ð Ñ IÒs s Ó IÒ 5 \Ó 5 IÒ\Ó Also, Z+<Òs ÓZ+<Ò 5 \ÓÐ 5 Ñ Z+<Ò\ÓÐ 5 Ñ & Þ Therefore, from Z+<Òs ÓÒ,3+= s Ð ÑÓ, we get 5 Ð Ñ 5 & Ð 5 Ñ, from which 5 &, and 5& Answer: D 2brove@rogerscom sam@utstattorontoedu

9 17 There is 8 sample value B, so BB, and the Buhlmann credibility estimate for the number of claims in Year 2 is ^B Ð ^Ñ, where + The model distribution \l has a geometric distribution and the parameter has a Pareto distribution with parameters and The hypothetical mean is IÒ\l Ó and the process variance is Z +<Ò\l Ó Ð Ñ Then IÒIÒ\l ÓÓIÒ Ó, +Z+<ÒIÒ\l ÓZ+<Ò ÓIÒ ÓÐIÒ ÓÑ ÐÑÐÑ Ò Ó ÐÑ ÐÑ ÐÑ ÓÓ IÒ Ð ÑÓIÒ ÓIÒ Ó ÐÑÐÑ ÐÑÐÑ ÐÑÐÑ Then + Ò ÐÑÐÑ Ó Ò ÐÑ ÐÑ Óß so that Þ + The Buhlmann credibility premium is B ^B Ð ^Ñ B Ð ÑÐ Ñ Answer: D 18 Survival in the context of this problem means not having had an accident The baseline survival distribution is parametric with a constant hazard rate A constant hazard rate corresponds to an exponential distribution, therefore baseline survival is exponential with a >Î mean of The pdf for baseline time of accident is 0Ð>Ñ! /, and the baseline survival >Î probability is WÐ>Ñ/ The hazard rate for individual 4with factors ^ and ^ is ^ ^! 2Ð>Ñ/ 4 2Ð>Ñ! The survival probability for that individual 4is / W4 Ð>Ñ ÒW! Ð>ÑÓ ^ ^, and the pdf for time of accident of that individual is 04 Ð>Ñ W4 w Ð>Ñ Ð/ ^ ^ ÑÒW! Ð>ÑÓ / ^ ^ 0! Ð>Ñ There are four individuals considered Individual 1 has ^!ß ^! and the accident is observed to occur at time 3 Since this is an observed accident time, in the likelihood function, we would include the density ^ ^ / ^ ^ 0 ÐÑ Ð/ ÑÒW Ð>ÑÓ 0 Ð>Ñ!! ÐÞÑÐ!ÑÐÞ!ÑÐ!Ñ ÐÞÑÐ!ÑÐÞ!ÑÐ!Ñ Î / Î Ð/ ÑÒ/ Ó / ; the log of that density is ÐÞÑÐ!ÑÐÞ!ÑÐ!Ñ ÐÞÑÐ!Ñ ÐÞ!ÑÐ!Ñ Ð/ ÑÐ Ñ 68ÐÑ Þ*% Individual 2 has ^!ß ^! and the accident occurs after time 6, so this is a right censored observation In the likelihood function we will include the survival probability WÐ'Ñ, and the log of that is ^ ^ ÐÞÑÐ!ÑÐÞ!ÑÐ!Ñ 68ÒW Ð'ÑÓ Ð/ Ñ 68ÒW! Ð'ÑÓ Ð/ ÑÐ Ñ Þ%&! ' 2brove@rogerscom sam@utstattorontoedu

10 18 continued Individual 3 has ^ ß ^! and the accident is observed to occur at time 7 Since this is an observed accident time, in the likelihood function, we would include the density ÐÞÑÐÑÐÞ!ÑÐ!Ñ ÐÞÑÐÑÐÞ!ÑÐ!Ñ (Î / (Î 0Ð(ÑÐ/ ÑÒ/ Ó / ; the log of that density is ÐÞÑÐÑÐÞ!ÑÐ!Ñ ( ( ÐÞÑÐÑ ÐÞ!ÑÐ!Ñ Ð/ ÑÐ Ñ 68ÐÑ Þ!(!& Individual 4 has ^ ß ^ %! and the accident occurs after time 8, so this is a right censored observation In the likelihood function we will include the survival probability WÐÑ %, and the log of that is ^ ^ ÐÞÑÐÑÐÞ!ÑÐ%!Ñ 68ÒW% ÐÑÓ Ð/ Ñ 68ÒW! ÐÑÓ Ð/ ÑÐ Ñ Þ( Þ The total loglikelihood function value is Þ*% Þ%&! Þ!( Þ( (Þ& Answer: C 19 A False See the bottom of page 427 of the Loss Models book B False The K-S test is applied to individual data C False See page 430 of the Loss Models text D False The critical value depends on the number of cells, the number of estimated parameters, and the level of significance Answer: E 20 The Buhlmann estimate is ^\ Ð ^Ñ _ In this problem we are given a single value of \(2 claims in year 1, so 8 and \ The hypothetical mean is IÒ\l Ó Ð!ÑÐ Ñ ÐÑÐ Ñ ÐÑÐ Ñ & The process variance is Z+<Ò\l ÓIÒ\ l ÓÐIÒ\l ÓÑ ÒÐ! ÑÐ Ñ Ð ÑÐ Ñ Ð ÑÐ ÑÓ Ð & Ñ * & Then IÒ IÒ\l Ó Ó IÒ & Ó &IÒ Ó &ÒÐÞ!&ÑÐÞÑ ÐÞÑÐÞÑÓ Þ&, + Z +<Ò IÒ\l Ó Ó Z +<Ò & Ó &Z +<Ò Ó &ÐÞ!& ÞÑ ÐÞÑÐÞÑ Þ&, ÓÓIÒ* & Ó*IÒ Ó&IÒ Ó *ÐÞÑ &ÒÐÞ!&Ñ ÐÞÑ ÐÞÑ ÐÞÑÓ Þ% Then ^ Þ%', and the Buhlmann credibility estimate is Þ% Þ& ÐÞ%'ÑÐÑ Ð Þ%'ÑÐÞ&Ñ Þ'* Answer: B 2brove@rogerscom sam@utstattorontoedu

11 21 The prior distribution is gamma with & and The model distribution is Poisson given - The combination of gamma prior distribution and Poisson model distribution results in a posterior distribution that is also gamma If there are 8 observed values of B, say B ß B ß ÞÞÞß B8, the posterior distribution has parameters w w D B3 and 8 We are given 8 observed values, B & and B w w Î Therefore, the posterior distribution has parameters &ß ÐÐÎÑÑ % The mean of the posterior distribution is w w ÐÑÐ % ÑÞ& Answer: B 22 The graph below illustrates the three kernel functions The first curve on the left is the kernel density function for the sample point C ; 5 ÐBÑ ÐB Ñ! Ÿ B Ÿ È 1 The middle curve is the kernel density function for the sample point C ; 5ÐBÑ È 1 ÐBÑ ŸBŸ% The curve on the right is the kernel density function for the sample point C & ; 5ÐBÑ & È 1 ÐB&Ñ %ŸBŸ' The empirical probabilities are :ÐÑ Þ& ß :ÐÑ Þ& (2 sample values at 3, and :Ð&Ñ Þ& The kernel density estimator is :ÐC Ñ5 ÐBÑ ÐÞ&Ñ5 ÐBÑ ÐÞ&Ñ5 ÐBÑ ÐÞ&Ñ5 ÐBÑÞ 4 4 C & 4 We see that the middle curve has double the coefficient as the curves on the left and right, so the middle curve is doubled, with resulting graph Answer: D 2brove@rogerscom sam@utstattorontoedu

12 23 A False See the middle paragraph on page 486 of the Loss Models book B False Cubic splines pass through all data points C False The stated requirement is for the natural spline D True See page 495 of the Loss Models book E False The inequality should be reversed See Theorem 155 on page 501 of the Loss Models book Answer: D 24 The first and second moments of the Pareto distribution are and ÐÑÐÑ The sample estimate of the first and second moments are!!&! & &!!! &! 2 and & 701,4016 ÐÑÐÑ Dividing then second expression by the square of the first expression results in 2 ÐÑ (!ß%!Þ' ÐÑÐÑ Š Ð&!Ñ Þ( Then solving for results in %Þ(' Solving for results in &!Ð Ñ *Þ The limited expected value with limit 500 is Þ(' *Þ *Þ IÒ\ &!!Ó Š Ò Š &!! Ó Š Þ(' Ò Š &!!*Þ Ó *' Answer: C 2 25 The credibility factor is ^s 7, where 7! and s+ is given as Þ 7@s +s 83 s@ Ð8 ÑÐ8 Ñ 734Ð\ 34 \Ñ 3 Then ÐÑÐÑ 3 4 Ò&!Ð!! (Þ(Ñ '!Ð&! (Þ(Ñ!!Ð'! (Þ*&Ñ *!Ð!! (Þ*&Ñ Ó (ß *&Þ' ^s! Þ%** Þ Answer: B! (ß*&Þ' '&Þ! 2brove@rogerscom sam@utstattorontoedu

13 26 The ogive is the cdf of the data based on linear interpolation between interval endpoints From the data set, 16 of 100 claims are below 1,000, so the empirical cdf at 1,000 is J!! Ðß!!!Ñ Þ' Similarly, J!! Ðß!!!Ñ Þ ß J!! Ð&ß!!!Ñ Þ' ß J!! Ð!ß!!!Ñ Þß etc We wish to find T Òß!!! \ Ÿ 'ß!!!Ó We implicitly assume that losses are uniformly distributed within each interval (this implies the linearity of J ÐBÑ within each interval, so that the estimated probability is!! J Ð'ß!!!Ñ J Ðß!!!Ñ Þ''' Þ(! Þ*'!!!! Answer: B 27 The pdf for the Pareto with!ß!!! is 0ÐBÑ!ß!!! ÐB!ß!!!Ñ!ß!!!!ß!!! and the cdf is JÐBÑ Š B!ß!!!, so that JÐBÑ Š B!ß!!! Þ The log of the density is 68 0ÐBÑ 68 68!ß!!! Ð Ñ 68ÐB!ß!!!Ñ,!ß!!! and the log of J ÐBÑ is 68Ð J ÐBÑÑ 68Š B!ß!!! The derivative with respect to of 68 0ÐBÑ is 68 0ÐBÑ 68!ß!!! 68ÐB!ß!!!Ñ,!ß!!! and the derivative with respect to of 68Ð J ÐBÑÑ is 68Š B!ß!!! For each of the 3 losses of amount 750 with deductible 200, the likelihood function has a factor of the conditional density 0Ð(&!l\!!Ñ 0Ð(&!Ñ J Ð!!Ñ, and the derivative with respect to of the log of the conditional density is!ß!!! Ò68 0Ð(&!Ñ 68Ð J Ð!!ÑÑÓ 68!ß!!! 68Ð!ß (&!Ñ 68Š!ß!! 68!ß!! 68Ð!ß (&!Ñ For each of the 3 losses of amount 200 with no deductible (and policy limit 10,000, the likelihood function has a factor of 0Ð!!Ñ, and the derivative of the log is 68!ß!!! 68Ð!ß!!Ñ For each of the 4 losses of amount 300 with no deductible (and policy limit 20,000, the likelihood function has a factor of 0Ð!!Ñ, and the derivative of the log is 68!ß!!! 68Ð!ß!!Ñ For each of the 6 losses above 10,000 with non deductible and policy limit 10,000, the likelihood!ß!!! factor is J Ð!ß!!!Ñ, and the derivative of the log of the likelihood factor is 68Š!ß!!! For each of the 4 losses of amount 400 with deductible 300, the likelihood function has a factor of the conditional density 0Ð%!!l\!!Ñ 0Ð%!!Ñ J Ð!!Ñ, and the derivative with respect to of the log of the conditional density is!ß!!! Ò68 0Ð%!!Ñ 68Ð J Ð!!ÑÑÓ 68!ß!!! 68Ð!ß %!!Ñ 68Š!ß!! 68!ß!! 68Ð!ß %!!Ñ 2brove@rogerscom sam@utstattorontoedu

14 27 continued The derivative of the log likelihood is the sum of all the derivatives listed above, and it is set equal to 0 to solve for the mle of : 68!ß!! 68!ß (&! 68!ß!!! 68Ð!ß!!Ñ % 68!ß!!! 68Ð!ß!!Ñ ' 68!ß!!! 68Ð!ß!!!Ñ % 68!ß!! 68Ð!ß %!!Ñ % 68!ß!!! 68!ß (&! ' 68!ß!!! % 68!ß %!!! Solving for results in Þ!* Answer: C 28 Since the information provided is the number of claims per policyholder over a 2-year period, we will use \ to denote the number of claims in a two year period Since the number of claims per year for an individual has a Poisson distribution, the number of claims in a 2-year period will also have a Poisson distribution Each policyholder has a mean, say, for the expected number of claims in 2 years, where varies from policyholder to another The 100 observations vary over different values of Using the semiparametric nonempirical Bayes approach, we have &!Ð!Ñ!ÐÑ&ÐÑ%ÐÑÐ%Ñ s!! Þ(' and s+!! Ò D\ 3!!\ Ó \ ** Ò%!!ÐÞ('Ñ Ó Þ(' Þ!*!* For a policyholder who has had 1 claim over the 2-year period, we have 1 observation (since the random variable \ refers to the number of claims in a 2-year period, we have information for one 2-year period Therefore the credibility factor for this policy holder is ^ Þ!' Þ(' The estimate for the expected number of claims in the next 2-year period Þ!*!* for a policyholder who had 1 in claim in the first 2-year period is ÐÞ!'ÑÐÑ Ð Þ!'ÑÐÞ('Ñ Þ(' The estimate for the expected number of claims in the Year 3 (the first half of the next 2-year period is ÐÞ('Ñ Þ* Þ Answer: C 29 The baseline hazard rate is parametric so we can formulate a full (no partial likelihood > function We find LÐ>Ñ! ' 2ÐBÑB >!! ÞThen for baseline survival, the survival LÐ>Ñ! >Î w > >Î probability is W! Ð>Ñ / /, and the density is 0! Ð>Ñ W! Ð>Ñ / Baseline is loss amount distribution for Class A For Class B, B/ > / > / Î the hazard rate is 2F ÐBÑ / 2! ÐBÑ, and LFÐ>Ñ ß WFÐ>Ñ /, >/ and 0 Ð>Ñ > / / Î F The likelihood function for the given data, with Class A losses of 1 and 3 and Class B losses of 2 and 4 is 2brove@rogerscom sam@utstattorontoedu

15 29 continued ' %/ / P 0! ÐÑ 0! ÐÑ 0FÐÑ 0FÐ%Ñ / / / / % / Þ % Ð!!/ ÑÎ Î *Î %/ Î '/ Î!!/ The loglikelihood is 68 P 68 % %68 Taking partial derivatives and setting equal to 0, we get ` %!!/ `!/ ` 68 P! ` 68 P and! From the second equation we have!/, and then substituting this into the first equation, %!!/ we have!!/!!/ Solving for / results in /, so that 68 Þ'* Þ Answer: A 30 Notice that at the three knots we have 2ÐÑ ß2Ð!Ñ!ß2ÐÑ, so the three data points actually lie in a straight line The natural cubic spline is the spline whose 2nd derivative is 0 at the end knots But the straight line through the three points has a 2nd derivative of 0 (any linear function is of the form C7B5 and has 2nd derivative 0 Therefore, the natural spline is simply the straight line through the knots The 2nd derivative of this spline is 0 at all points We can also find the spline using the algebraic approach The natural spline on the interval Òß!Ó is of the form ww 0! ÐBÑ+!,! ÐBÑ-! ÐBÑ! ÐBÑ, and 0! ÐBÑ-! '! ÐBÑ For a natural spline one 3 knots, we have 7! 7! and CC CC!!!ÐÑ Ð2! 2 Ñ7 ' Š 2 2, so that Ð Ñ7 ' Š! Solving for 7 results in 7! We use the relationships -4 and 4 '24 to get -!! and!! Therefore, 0! ww Ð!Þ&Ñ! Answer: C Þ ÐBÎ Ñ Þ 31 The pdf of this Weibull with 7 Þ is 0ÐBÑ ÞB / Þ, its log is B 68 0ÐBÑ 68 Þ Þ 68 B Þ 68 Ð Ñ Þ, and the derivative of 68 0ÐBÑ with respect to is Þ ÞB 68 0ÐBÑ Þ Þ Þ B Þ ÞB The derivative of 68Ð J ÐBÑÑ is 68Ð J ÐBÑÑ Ò Ð Ñ Ó Þ For the given data points, the derivative of the loglikelihood function is the sum of the derivatives of 68 0ÐBÑ for the four given claim amounts plus 2 the derivative of 68Ò J Ð!!!ÑÓ 2brove@rogerscom sam@utstattorontoedu

16 31 continued Þ Þ Þ Þ! Þ Þ %! Þ 68 P Þ Þ Þ Þ Þ Þ!! Þ Þ &%! Þ!!! Þ Þ Þ! (note that the factor 2 cancels from all terms Solving for results in the mle estimate ' An alternative solution makes use of the relation between the Weibull and exponential 7 distributions If \ has a Weibull distribution with 7 known and unknown, then ^ \ has an exponential distribution with mean 7 If B ßßßÞB8 are sample values from a Weibull distribution with 7 known, then D B ß ÞÞÞß D B behave like a sample from an exponential distribution with mean 7 (this applies to known B-amounts and limit amounts also Þ Þ Þ Þ Þ We are given 7 Þ, so the values! ß %! ß!! ß &%! and the two limit values of!!! are from an exponential distribution with mean Þ When a sample from an exponential distribution has uncensored values and censored (limit values, the mle of the exponential mean sum of all observed values (including limit values is number of uncensored values In this case, the mle of Þ is Þ Þ Þ Þ Þ! %!!! &%!!!! % &Þ!', so that the mle of is Ð&Þ!'Ñ & ( Answer: E 32 The Buhlmann estimate is ^\ Ð ^Ñ, where \ is the number of claims in the first year This is a linear function of \, so answer E can be eliminated, since the Buhlmann estimate % is not linear The Bayesian estimate is IÒ\ l\ 8Ó ' - 1Ð-l\ 8Ñ - The Bayesian estimate must be between 1 and 4, since that is the interval on which - is defined This eliminates answers B and D, since B has a Bayesian estimate greater than 4 and D has a Bayesian estimate less than 1 The following reasoning can be used to eliminate answer C If the Buhlmann estimate for the 2nd year is 5 (as in graph C when the number of claims in the Þ& Þ& first year is 0, then Ð ^Ñ Þ& Since Ÿ Ÿ %, it follows that % Ÿ Ÿ Þ&, and then Þ& since ^, we have Þ&Ÿ^ŸÞ(& But then, if \* in the first year, the Buhlmann estimate for the 2nd year will be *^ Ð ^Ñ %Þ&, which is inconsistent with the Buhlmann estimate in graph C when \* in the first year Answer: A 2brove@rogerscom sam@utstattorontoedu

17 33 The hypothesized probabilities for the intervals are J ÐÑ Þ!& ß J Ð&Ñ Þ! ß J Ð(Ñ Þ'! ß J ÐÑ Þ! ß J Ð Ñ Therefore, T Ò\ Ÿ Ó Þ!& ß T Ò \ Ÿ &Ó Þ!*&ß T Ò& \ Ÿ (Ó Þ&!! ß T Ò( \ Ÿ Ó Þ!! ß T Ò \Ó Þ(! For the 300 observations, the expected numbers for each interval, based on the hypothesized distribution are Interval Expected Number of Observations Actual Number of Observations Ò! ß Ó!!ÐÞ!&Ñ!Þ& & Ð ß &Ó!!ÐÞ!*&Ñ Þ& % Ð& ß (Ó!!ÐÞ&!!Ñ &! ( Ð( ß Ó!!ÐÞ!!Ñ '! '' Ð ß Ñ!!ÐÞ(!Ñ & &! The chi-square statistic is Ð&!Þ&Ñ Ð%Þ&Ñ Ð(&!Ñ Ð'''!Ñ Ð&&!Ñ U!Þ& Þ& &! '! & Þ! The degrees of freedom in this chi-square test is &%(there is no estimation done, so no degrees of freedom are lost due to estimation From the chi-square table with 4 degrees of freedom, we have ; Þ!& Ð%Ñ *Þ% (95-th percentile and ; Þ!& Ð%Ñ Þ% (975-percentile The null hypothesis that the given data comes from the known distribution is rejected at the 5% level of significance because U Þ! *Þ%, but the hypothesis is no rejected at the 25% level of significance Answer: C 68 B 34 The cdf of the lognormal distribution is JÐBÑ FŠ 5 According to the inversion 68 B method, given uniform number?, we solve for B from? JÐBÑ FŠ 5 68 B &Þ'? Þ'(* FŠ Þ(& ; from the standard normal distribution table, we see that 68 B &Þ' FÐÞÑ Þ'(* ; therefore, Þ(& Þ, so that the simulated value of B is 33866? Þ%'! Þ&* and Þ&* FÐÞÑ, so that Þ%'! FÐ ÞÑ, and then 68 B &Þ' Þ(& Þ and the simulated value of B is 25089? Þ*%& F ÐÞ'Ñ p B *(Þ&, so the policy limit of 400 is applied? Þ!! F Ð Þ'%Ñ p B *%Þ'? Þ( F ÐÞÑ p B %*Þ(&, and the policy limit of 400 is applied? Þ%!( F Ð ÞÑ p B Þ(' Þ''&!Þ*%!!*%Þ'%!!Þ(' The average payment per claim is ' ' Answer: A 2brove@rogerscom sam@utstattorontoedu

18 35 We wish to find IÒ\ l\ Ó Using the Bayesian methodology, this can be formulated as IÒ\ l Ó T Ò l\ Ó IÒ\ l &Ó T Ò &l\ Ó The mean of the geometric distribution with parameter is, so IÒ\l Ó and IÒ\l &Ó& The conditional probabilities can be found in the following way: TÐ Ñ TÐ &Ñ % & & TÒ\ l Ó ÐÑ ( TÒ\ l &Ó Ð&Ñ ' Ì Ì TÒ\ Ó TÒ\ &Ó % &! T Ò\ l Ó T Ð Ñ T Ò\ l &Ó T Ð &Ñ '% Ì % &! T Ò\ Ó T Ò\ Ó T Ò\ &Ó '% Þ'&% Ì T Ò\ Ó %Î TÒ l\ Ó T Ò\ Ó Þ'&% Þ*! and T Ò\ &Ó &!Î'% TÒ l\ &Ó T Ò\ Ó Þ'&% Þ'! Then, Answer: C IÒ\ l\ Ó ÐÑÐÞ*Ñ Ð&ÑÐÞ'Ñ Þ 2brove@rogerscom sam@utstattorontoedu

SIMULATION - PROBLEM SET 1

SIMULATION - PROBLEM SET 1 SIMULATION - PROBLEM SET 1 " if! Ÿ B Ÿ 1. The random variable X has probability density function 0ÐBÑ œ " $ if Ÿ B Ÿ.! otherwise Using the inverse transform method of simulation, find the random observation

More information

NOVEMBER 2005 EXAM NOVEMBER 2005 CAS COURSE 3 EXAM SOLUTIONS 1. Maximum likelihood estimation: The log of the density is 68 0ÐBÑ œ 68 ) Ð) "Ñ 68 B. The loglikelihood function is & j œ 68 0ÐB Ñ œ & 68 )

More information

Exam C Solutions Spring 2005

Exam C Solutions Spring 2005 Exam C Solutions Spring 005 Question # The CDF is F( x) = 4 ( + x) Observation (x) F(x) compare to: Maximum difference 0. 0.58 0, 0. 0.58 0.7 0.880 0., 0.4 0.680 0.9 0.93 0.4, 0.6 0.53. 0.949 0.6, 0.8

More information

SPRING 2007 EXAM C SOLUTIONS

SPRING 2007 EXAM C SOLUTIONS SPRING 007 EXAM C SOLUTIONS Question #1 The data are already shifted (have had the policy limit and the deductible of 50 applied). The two 350 payments are censored. Thus the likelihood function is L =

More information

Practice Exam 1. (A) (B) (C) (D) (E) You are given the following data on loss sizes:

Practice Exam 1. (A) (B) (C) (D) (E) You are given the following data on loss sizes: Practice Exam 1 1. Losses for an insurance coverage have the following cumulative distribution function: F(0) = 0 F(1,000) = 0.2 F(5,000) = 0.4 F(10,000) = 0.9 F(100,000) = 1 with linear interpolation

More information

DEPARTMENT OF MATHEMATICS AND STATISTICS QUALIFYING EXAMINATION II IN MATHEMATICAL STATISTICS SATURDAY, SEPTEMBER 24, 2016

DEPARTMENT OF MATHEMATICS AND STATISTICS QUALIFYING EXAMINATION II IN MATHEMATICAL STATISTICS SATURDAY, SEPTEMBER 24, 2016 DEPARTMENT OF MATHEMATICS AND STATISTICS QUALIFYING EXAMINATION II IN MATHEMATICAL STATISTICS SATURDAY, SEPTEMBER 24, 2016 Examiners: Drs. K. M. Ramachandran and G. S. Ladde INSTRUCTIONS: a. The quality

More information

QUALIFYING EXAMINATION II IN MATHEMATICAL STATISTICS SATURDAY, MAY 9, Examiners: Drs. K. M. Ramachandran and G. S. Ladde

QUALIFYING EXAMINATION II IN MATHEMATICAL STATISTICS SATURDAY, MAY 9, Examiners: Drs. K. M. Ramachandran and G. S. Ladde QUALIFYING EXAMINATION II IN MATHEMATICAL STATISTICS SATURDAY, MAY 9, 2015 Examiners: Drs. K. M. Ramachandran and G. S. Ladde INSTRUCTIONS: a. The quality is more important than the quantity. b. Attempt

More information

ACTEX CAS EXAM 3 STUDY GUIDE FOR MATHEMATICAL STATISTICS

ACTEX CAS EXAM 3 STUDY GUIDE FOR MATHEMATICAL STATISTICS ACTEX CAS EXAM 3 STUDY GUIDE FOR MATHEMATICAL STATISTICS TABLE OF CONTENTS INTRODUCTORY NOTE NOTES AND PROBLEM SETS Section 1 - Point Estimation 1 Problem Set 1 15 Section 2 - Confidence Intervals and

More information

ACT455H1S - TEST 2 - MARCH 25, 2008

ACT455H1S - TEST 2 - MARCH 25, 2008 ACT455H1S - TEST 2 - MARCH 25, 2008 Write name and student number on each page. Write your solution for each question in the space provided. For the multiple decrement questions, it is always assumed that

More information

PART I. Multiple choice. 1. Find the slope of the line shown here. 2. Find the slope of the line with equation $ÐB CÑœ(B &.

PART I. Multiple choice. 1. Find the slope of the line shown here. 2. Find the slope of the line with equation $ÐB CÑœ(B &. Math 1301 - College Algebra Final Exam Review Sheet Version X This review, while fairly comprehensive, should not be the only material used to study for the final exam. It should not be considered a preview

More information

Dr. H. Joseph Straight SUNY Fredonia Smokin' Joe's Catalog of Groups: Direct Products and Semi-direct Products

Dr. H. Joseph Straight SUNY Fredonia Smokin' Joe's Catalog of Groups: Direct Products and Semi-direct Products Dr. H. Joseph Straight SUNY Fredonia Smokin' Joe's Catalog of Groups: Direct Products and Semi-direct Products One of the fundamental problems in group theory is to catalog all the groups of some given

More information

EXCERPTS FROM ACTEX CALCULUS REVIEW MANUAL

EXCERPTS FROM ACTEX CALCULUS REVIEW MANUAL EXCERPTS FROM ACTEX CALCULUS REVIEW MANUAL Table of Contents Introductory Comments SECTION 6 - Differentiation PROLEM SET 6 TALE OF CONTENTS INTRODUCTORY COMMENTS Section 1 Set Theory 1 Section 2 Intervals,

More information

Parameter Estimation

Parameter Estimation Parameter Estimation Chapters 13-15 Stat 477 - Loss Models Chapters 13-15 (Stat 477) Parameter Estimation Brian Hartman - BYU 1 / 23 Methods for parameter estimation Methods for parameter estimation Methods

More information

Inner Product Spaces

Inner Product Spaces Inner Product Spaces In 8 X, we defined an inner product? @? @?@ ÞÞÞ? 8@ 8. Another notation sometimes used is? @? ß@. The inner product in 8 has several important properties ( see Theorem, p. 33) that

More information

SIMULTANEOUS CONFIDENCE BANDS FOR THE PTH PERCENTILE AND THE MEAN LIFETIME IN EXPONENTIAL AND WEIBULL REGRESSION MODELS. Ping Sa and S.J.

SIMULTANEOUS CONFIDENCE BANDS FOR THE PTH PERCENTILE AND THE MEAN LIFETIME IN EXPONENTIAL AND WEIBULL REGRESSION MODELS. Ping Sa and S.J. SIMULTANEOUS CONFIDENCE BANDS FOR THE PTH PERCENTILE AND THE MEAN LIFETIME IN EXPONENTIAL AND WEIBULL REGRESSION MODELS " # Ping Sa and S.J. Lee " Dept. of Mathematics and Statistics, U. of North Florida,

More information

Math 131, Exam 2 Solutions, Fall 2010

Math 131, Exam 2 Solutions, Fall 2010 Math 131, Exam 2 Solutions, Fall 2010 Part I consists of 14 multiple choice questions (orth 5 points each) and 5 true/false questions (orth 1 point each), for a total of 75 points. Mark the correct anser

More information

Math 131 Exam 4 (Final Exam) F04M

Math 131 Exam 4 (Final Exam) F04M Math 3 Exam 4 (Final Exam) F04M3.4. Name ID Number The exam consists of 8 multiple choice questions (5 points each) and 0 true/false questions ( point each), for a total of 00 points. Mark the correct

More information

Nonparametric Model Construction

Nonparametric Model Construction Nonparametric Model Construction Chapters 4 and 12 Stat 477 - Loss Models Chapters 4 and 12 (Stat 477) Nonparametric Model Construction Brian Hartman - BYU 1 / 28 Types of data Types of data For non-life

More information

Estimation for Modified Data

Estimation for Modified Data Definition. Estimation for Modified Data 1. Empirical distribution for complete individual data (section 11.) An observation X is truncated from below ( left truncated) at d if when it is at or below d

More information

Solutions to the Spring 2015 CAS Exam ST

Solutions to the Spring 2015 CAS Exam ST Solutions to the Spring 2015 CAS Exam ST (updated to include the CAS Final Answer Key of July 15) There were 25 questions in total, of equal value, on this 2.5 hour exam. There was a 10 minute reading

More information

TABLE OF CONTENTS CHAPTER 1 COMBINATORIAL PROBABILITY 1

TABLE OF CONTENTS CHAPTER 1 COMBINATORIAL PROBABILITY 1 TABLE OF CONTENTS CHAPTER 1 COMBINATORIAL PROBABILITY 1 1.1 The Probability Model...1 1.2 Finite Discrete Models with Equally Likely Outcomes...5 1.2.1 Tree Diagrams...6 1.2.2 The Multiplication Principle...8

More information

Distribution Fitting (Censored Data)

Distribution Fitting (Censored Data) Distribution Fitting (Censored Data) Summary... 1 Data Input... 2 Analysis Summary... 3 Analysis Options... 4 Goodness-of-Fit Tests... 6 Frequency Histogram... 8 Comparison of Alternative Distributions...

More information

CREDIBILITY - PROBLEM SET 3 Bayesian Credibility - Discrete Prior

CREDIBILITY - PROBLEM SET 3 Bayesian Credibility - Discrete Prior CREDIBILITY - PROBLEM SET 3 Bayesian Credibility - Discrete Prior 1 Two bowls each contain 10 similarly shaped balls Bowl 1 contains 5 red and 5 white balls (equally likely to be chosen Bowl 2 contains

More information

SVM example: cancer classification Support Vector Machines

SVM example: cancer classification Support Vector Machines SVM example: cancer classification Support Vector Machines 1. Cancer genomics: TCGA The cancer genome atlas (TCGA) will provide high-quality cancer data for large scale analysis by many groups: SVM example:

More information

Course 4 Solutions November 2001 Exams

Course 4 Solutions November 2001 Exams Course 4 Solutions November 001 Exams November, 001 Society of Actuaries Question #1 From the Yule-Walker equations: ρ φ + ρφ 1 1 1. 1 1+ ρ ρφ φ Substituting the given quantities yields: 0.53 φ + 0.53φ

More information

Stat 643 Problems. a) Show that V is a sub 5-algebra of T. b) Let \ be an integrable random variable. Find EÐ\lVÑÞ

Stat 643 Problems. a) Show that V is a sub 5-algebra of T. b) Let \ be an integrable random variable. Find EÐ\lVÑÞ Stat 643 Problems 1. Let ( HT, ) be a measurable space. Suppose that., T and Uare 5-finite positive measures on this space with T Uand U...T.T.U a) Show that T. and œ a.s......u.....u.u Š.. b) Show that

More information

Inner Product Spaces. Another notation sometimes used is X.?

Inner Product Spaces. Another notation sometimes used is X.? Inner Product Spaces X In 8 we have an inner product? @? @?@ ÞÞÞ? 8@ 8Þ Another notation sometimes used is X? ß@? @? @? @ ÞÞÞ? @ 8 8 The inner product in?@ ß in 8 has several essential properties ( see

More information

Gene expression experiments. Infinite Dimensional Vector Spaces. 1. Motivation: Statistical machine learning and reproducing kernel Hilbert Spaces

Gene expression experiments. Infinite Dimensional Vector Spaces. 1. Motivation: Statistical machine learning and reproducing kernel Hilbert Spaces MA 751 Part 3 Gene expression experiments Infinite Dimensional Vector Spaces 1. Motivation: Statistical machine learning and reproducing kernel Hilbert Spaces Gene expression experiments Question: Gene

More information

MATH 1301 (College Algebra) - Final Exam Review

MATH 1301 (College Algebra) - Final Exam Review MATH 1301 (College Algebra) - Final Exam Review This review is comprehensive but should not be the only material used to study for the final exam. It should not be considered a preview of the final exam.

More information

Practice Problems Section Problems

Practice Problems Section Problems Practice Problems Section 4-4-3 4-4 4-5 4-6 4-7 4-8 4-10 Supplemental Problems 4-1 to 4-9 4-13, 14, 15, 17, 19, 0 4-3, 34, 36, 38 4-47, 49, 5, 54, 55 4-59, 60, 63 4-66, 68, 69, 70, 74 4-79, 81, 84 4-85,

More information

Infinite Dimensional Vector Spaces. 1. Motivation: Statistical machine learning and reproducing kernel Hilbert Spaces

Infinite Dimensional Vector Spaces. 1. Motivation: Statistical machine learning and reproducing kernel Hilbert Spaces MA 751 Part 3 Infinite Dimensional Vector Spaces 1. Motivation: Statistical machine learning and reproducing kernel Hilbert Spaces Microarray experiment: Question: Gene expression - when is the DNA in

More information

Part I consists of 14 multiple choice questions (worth 5 points each) and 5 true/false question (worth 1 point each), for a total of 75 points.

Part I consists of 14 multiple choice questions (worth 5 points each) and 5 true/false question (worth 1 point each), for a total of 75 points. Math 131 Exam 1 Solutions Part I consists of 14 multiple choice questions (orth 5 points each) and 5 true/false question (orth 1 point each), for a total of 75 points. 1. The folloing table gives the number

More information

TABLE OF CONTENTS - VOLUME 1

TABLE OF CONTENTS - VOLUME 1 TABLE OF CONTENTS - VOLUME 1 INTRODUCTORY COMMENTS MODELING SECTION 1 - PROBABILITY REVIEW PROBLEM SET 1 LM-1 LM-9 SECTION 2 - REVIEW OF RANDOM VARIABLES - PART I PROBLEM SET 2 LM-19 LM-29 SECTION 3 -

More information

The Rational Numbers

The Rational Numbers The Rational Numbers Fields The system of integers that e formally defined is an improvement algebraically on the hole number system = (e can subtract in ) But still has some serious deficiencies: for

More information

It's Only Fitting. Fitting model to data parameterizing model estimating unknown parameters in the model

It's Only Fitting. Fitting model to data parameterizing model estimating unknown parameters in the model It's Only Fitting Fitting model to data parameterizing model estimating unknown parameters in the model Likelihood: an example Cohort of 8! individuals observe survivors at times >œ 1, 2, 3,..., : 8",

More information

Engineering Mathematics (E35 317) Final Exam December 18, 2007

Engineering Mathematics (E35 317) Final Exam December 18, 2007 Engineering Mathematics (E35 317) Final Exam December 18, 2007 This exam contains 18 multile-choice roblems orth to oints each, five short-anser roblems orth one oint each, and nine true-false roblems

More information

1. Classify each number. Choose all correct answers. b. È # : (i) natural number (ii) integer (iii) rational number (iv) real number

1. Classify each number. Choose all correct answers. b. È # : (i) natural number (ii) integer (iii) rational number (iv) real number Review for Placement Test To ypass Math 1301 College Algebra Department of Computer and Mathematical Sciences University of Houston-Downtown Revised: Fall 2009 PLEASE READ THE FOLLOWING CAREFULLY: 1. The

More information

You may have a simple scientific calculator to assist with arithmetic, but no graphing calculators are allowed on this exam.

You may have a simple scientific calculator to assist with arithmetic, but no graphing calculators are allowed on this exam. Math 131 Exam 3 Solutions You may have a simple scientific calculator to assist ith arithmetic, but no graphing calculators are alloed on this exam. Part I consists of 14 multiple choice questions (orth

More information

TOTAL DIFFERENTIALS. In Section B-4, we observed that.c is a pretty good approximation for? C, which is the increment of

TOTAL DIFFERENTIALS. In Section B-4, we observed that.c is a pretty good approximation for? C, which is the increment of TOTAL DIFFERENTIALS The differential was introduced in Section -4. Recall that the differential of a function œ0ðñis w. œ 0 ÐÑ.Þ Here. is the differential with respect to the independent variable and is

More information

Errata for the ASM Study Manual for Exam P, Fourth Edition By Dr. Krzysztof M. Ostaszewski, FSA, CFA, MAAA

Errata for the ASM Study Manual for Exam P, Fourth Edition By Dr. Krzysztof M. Ostaszewski, FSA, CFA, MAAA Errata for the ASM Study Manual for Exam P, Fourth Edition By Dr. Krzysztof M. Ostaszewski, FSA, CFA, MAAA (krzysio@krzysio.net) Effective July 5, 3, only the latest edition of this manual will have its

More information

Subject CT6. CMP Upgrade 2013/14. CMP Upgrade

Subject CT6. CMP Upgrade 2013/14. CMP Upgrade CT6: CMP Upgrade 013/14 Page 1 Subject CT6 CMP Upgrade 013/14 CMP Upgrade This CMP Upgrade lists all significant changes to the Core Reading and the ActEd material since last year so that you can manually

More information

Risk Models and Their Estimation

Risk Models and Their Estimation ACTEX Ac a d e m i c Se r i e s Risk Models and Their Estimation S t e p h e n G. K e l l i s o n, F S A, E A, M A A A U n i v e r s i t y o f C e n t r a l F l o r i d a ( R e t i r e d ) R i c h a r

More information

Math 132, Spring 2003 Exam 2: Solutions

Math 132, Spring 2003 Exam 2: Solutions Math 132, Spring 2003 Exam 2: Solutions No calculators with a CAS are allowed. e sure your calculator is set for radians, not degrees if you do any calculus computations with trig functions. Part I, Multiple

More information

Probability basics. Probability and Measure Theory. Probability = mathematical analysis

Probability basics. Probability and Measure Theory. Probability = mathematical analysis MA 751 Probability basics Probability and Measure Theory 1. Basics of probability 2 aspects of probability: Probability = mathematical analysis Probability = common sense Probability basics A set-up of

More information

Relations. Relations occur all the time in mathematics. For example, there are many relations between # and % À

Relations. Relations occur all the time in mathematics. For example, there are many relations between # and % À Relations Relations occur all the time in mathematics. For example, there are many relations between and % À Ÿ% % Á% l% Ÿ,, Á ß and l are examples of relations which might or might not hold between two

More information

Mixture distributions in Exams MLC/3L and C/4

Mixture distributions in Exams MLC/3L and C/4 Making sense of... Mixture distributions in Exams MLC/3L and C/4 James W. Daniel Jim Daniel s Actuarial Seminars www.actuarialseminars.com February 1, 2012 c Copyright 2012 by James W. Daniel; reproduction

More information

7. Random variables as observables. Definition 7. A random variable (function) X on a probability measure space is called an observable

7. Random variables as observables. Definition 7. A random variable (function) X on a probability measure space is called an observable 7. Random variables as observables Definition 7. A random variable (function) X on a probability measure space is called an observable Note all functions are assumed to be measurable henceforth. Note that

More information

Errata and updates for ASM Exam C/Exam 4 Manual (Sixth Edition) sorted by date

Errata and updates for ASM Exam C/Exam 4 Manual (Sixth Edition) sorted by date Errata for ASM Exam C/4 Study Manual (Sixth Edition) Sorted by Date 1 Errata and updates for ASM Exam C/Exam 4 Manual (Sixth Edition) sorted by date Please note that the SOA has announced that when using

More information

Math 1AA3/1ZB3 Sample Test 3, Version #1

Math 1AA3/1ZB3 Sample Test 3, Version #1 Math 1AA3/1ZB3 Sample Test 3, Version 1 Name: (Last Name) (First Name) Student Number: Tutorial Number: This test consists of 16 multiple choice questions worth 1 mark each (no part marks), and 1 question

More information

Contents. Preface to Second Edition Preface to First Edition Abbreviations PART I PRINCIPLES OF STATISTICAL THINKING AND ANALYSIS 1

Contents. Preface to Second Edition Preface to First Edition Abbreviations PART I PRINCIPLES OF STATISTICAL THINKING AND ANALYSIS 1 Contents Preface to Second Edition Preface to First Edition Abbreviations xv xvii xix PART I PRINCIPLES OF STATISTICAL THINKING AND ANALYSIS 1 1 The Role of Statistical Methods in Modern Industry and Services

More information

ASM Study Manual for Exam P, Second Edition By Dr. Krzysztof M. Ostaszewski, FSA, CFA, MAAA Errata

ASM Study Manual for Exam P, Second Edition By Dr. Krzysztof M. Ostaszewski, FSA, CFA, MAAA Errata ASM Study Manual for Exam P, Second Edition By Dr. Krzysztof M. Ostaszewski, FSA, CFA, MAAA (krzysio@krzysio.net) Errata Effective July 5, 3, only the latest edition of this manual will have its errata

More information

Dr. Nestler - Math 2 - Ch 3: Polynomial and Rational Functions

Dr. Nestler - Math 2 - Ch 3: Polynomial and Rational Functions Dr. Nestler - Math 2 - Ch 3: Polynomial and Rational Functions 3.1 - Polynomial Functions We have studied linear functions and quadratic functions Defn. A monomial or power function is a function of the

More information

Proofs Involving Quantifiers. Proof Let B be an arbitrary member Proof Somehow show that there is a value

Proofs Involving Quantifiers. Proof Let B be an arbitrary member Proof Somehow show that there is a value Proofs Involving Quantifiers For a given universe Y : Theorem ÐaBÑ T ÐBÑ Theorem ÐbBÑ T ÐBÑ Proof Let B be an arbitrary member Proof Somehow show that there is a value of Y. Call it B œ +, say Þ ÐYou can

More information

Errata and updates for ASM Exam C/Exam 4 Manual (Sixth Edition) sorted by page

Errata and updates for ASM Exam C/Exam 4 Manual (Sixth Edition) sorted by page Errata for ASM Exam C/4 Study Manual (Sixth Edition) Sorted by Page Errata and updates for ASM Exam C/Exam 4 Manual (Sixth Edition) sorted by page Please note that the SOA has announced that when using

More information

ASM Study Manual for Exam P, First Edition By Dr. Krzysztof M. Ostaszewski, FSA, CFA, MAAA Errata

ASM Study Manual for Exam P, First Edition By Dr. Krzysztof M. Ostaszewski, FSA, CFA, MAAA Errata ASM Study Manual for Exam P, First Edition By Dr. Krzysztof M. Ostaszewski, FSA, CFA, MAAA (krzysio@krzysio.net) Errata Effective July 5, 3, only the latest edition of this manual will have its errata

More information

8œ! This theorem is justified by repeating the process developed for a Taylor polynomial an infinite number of times.

8œ! This theorem is justified by repeating the process developed for a Taylor polynomial an infinite number of times. Taylor and Maclaurin Series We can use the same process we used to find a Taylor or Maclaurin polynomial to find a power series for a particular function as long as the function has infinitely many derivatives.

More information

Théorie Analytique des Probabilités

Théorie Analytique des Probabilités Théorie Analytique des Probabilités Pierre Simon Laplace Book II 5 9. pp. 203 228 5. An urn being supposed to contain the number B of balls, e dra from it a part or the totality, and e ask the probability

More information

e) D œ < f) D œ < b) A parametric equation for the line passing through Ð %, &,) ) and (#,(, %Ñ.

e) D œ < f) D œ < b) A parametric equation for the line passing through Ð %, &,) ) and (#,(, %Ñ. Page 1 Calculus III : Bonus Problems: Set 1 Grade /42 Name Due at Exam 1 6/29/2018 1. (2 points) Give the equations for the following geometrical objects : a) A sphere of radius & centered at the point

More information

Preface Introduction to Statistics and Data Analysis Overview: Statistical Inference, Samples, Populations, and Experimental Design The Role of

Preface Introduction to Statistics and Data Analysis Overview: Statistical Inference, Samples, Populations, and Experimental Design The Role of Preface Introduction to Statistics and Data Analysis Overview: Statistical Inference, Samples, Populations, and Experimental Design The Role of Probability Sampling Procedures Collection of Data Measures

More information

Solving SVM: Quadratic Programming

Solving SVM: Quadratic Programming MA 751 Part 7 Solving SVM: Quadratic Programming 1. Quadratic programming (QP): Introducing Lagrange multipliers α 4 and. 4 (can be justified in QP for inequality as well as equality constraints) we define

More information

Stat 5101 Lecture Notes

Stat 5101 Lecture Notes Stat 5101 Lecture Notes Charles J. Geyer Copyright 1998, 1999, 2000, 2001 by Charles J. Geyer May 7, 2001 ii Stat 5101 (Geyer) Course Notes Contents 1 Random Variables and Change of Variables 1 1.1 Random

More information

Math 1M03 (Version 1) Sample Exam

Math 1M03 (Version 1) Sample Exam Math 1M03 (Version 1) Sample Exam Name: (Last Name) (First Name) Student Number: Day Class Duration: 3 Hours Maximum Mark: 40 McMaster University Sample Final Examination This examination paper consists

More information

4 Testing Hypotheses. 4.1 Tests in the regression setting. 4.2 Non-parametric testing of survival between groups

4 Testing Hypotheses. 4.1 Tests in the regression setting. 4.2 Non-parametric testing of survival between groups 4 Testing Hypotheses The next lectures will look at tests, some in an actuarial setting, and in the last subsection we will also consider tests applied to graduation 4 Tests in the regression setting )

More information

Subject CS1 Actuarial Statistics 1 Core Principles

Subject CS1 Actuarial Statistics 1 Core Principles Institute of Actuaries of India Subject CS1 Actuarial Statistics 1 Core Principles For 2019 Examinations Aim The aim of the Actuarial Statistics 1 subject is to provide a grounding in mathematical and

More information

Statistical machine learning and kernel methods. Primary references: John Shawe-Taylor and Nello Cristianini, Kernel Methods for Pattern Analysis

Statistical machine learning and kernel methods. Primary references: John Shawe-Taylor and Nello Cristianini, Kernel Methods for Pattern Analysis Part 5 (MA 751) Statistical machine learning and kernel methods Primary references: John Shawe-Taylor and Nello Cristianini, Kernel Methods for Pattern Analysis Christopher Burges, A tutorial on support

More information

The Spotted Owls 5 " 5. = 5 " œ!þ")45 Ð'!% leave nest, 30% of those succeed) + 5 " œ!þ("= 5!Þ*%+ 5 ( adults live about 20 yrs)

The Spotted Owls 5  5. = 5  œ!þ)45 Ð'!% leave nest, 30% of those succeed) + 5  œ!þ(= 5!Þ*%+ 5 ( adults live about 20 yrs) The Spotted Owls Be sure to read the example at the introduction to Chapter in the textbook. It gives more general background information for this example. The example leads to a dynamical system which

More information

Engineering Mathematics (E35 317) Final Exam December 15, 2006

Engineering Mathematics (E35 317) Final Exam December 15, 2006 Engineering Mathematics (E35 317) Final Exam December 15, 2006 This exam contains six free-resonse roblems orth 36 oints altogether, eight short-anser roblems orth one oint each, seven multile-choice roblems

More information

Estimation of Quantiles

Estimation of Quantiles 9 Estimation of Quantiles The notion of quantiles was introduced in Section 3.2: recall that a quantile x α for an r.v. X is a constant such that P(X x α )=1 α. (9.1) In this chapter we examine quantiles

More information

This does not cover everything on the final. Look at the posted practice problems for other topics.

This does not cover everything on the final. Look at the posted practice problems for other topics. Class 7: Review Problems for Final Exam 8.5 Spring 7 This does not cover everything on the final. Look at the posted practice problems for other topics. To save time in class: set up, but do not carry

More information

Table of z values and probabilities for the standard normal distribution. z is the first column plus the top row. Each cell shows P(X z).

Table of z values and probabilities for the standard normal distribution. z is the first column plus the top row. Each cell shows P(X z). Table of z values and probabilities for the standard normal distribution. z is the first column plus the top row. Each cell shows P(X z). For example P(X.04) =.8508. For z < 0 subtract the value from,

More information

Suggestions - Problem Set (a) Show the discriminant condition (1) takes the form. ln ln, # # R R

Suggestions - Problem Set (a) Show the discriminant condition (1) takes the form. ln ln, # # R R Suggetion - Problem Set 3 4.2 (a) Show the dicriminant condition (1) take the form x D Ð.. Ñ. D.. D. ln ln, a deired. We then replace the quantitie. 3ß D3 by their etimate to get the proper form for thi

More information

STAT 2122 Homework # 3 Solutions Fall 2018 Instr. Sonin

STAT 2122 Homework # 3 Solutions Fall 2018 Instr. Sonin STAT 2122 Homeork Solutions Fall 2018 Instr. Sonin Due Friday, October 5 NAME (25 + 5 points) Sho all ork on problems! (4) 1. In the... Lotto, you may pick six different numbers from the set Ö"ß ß $ß ÞÞÞß

More information

Probability Distributions Columns (a) through (d)

Probability Distributions Columns (a) through (d) Discrete Probability Distributions Columns (a) through (d) Probability Mass Distribution Description Notes Notation or Density Function --------------------(PMF or PDF)-------------------- (a) (b) (c)

More information

Department of Statistical Science FIRST YEAR EXAM - SPRING 2017

Department of Statistical Science FIRST YEAR EXAM - SPRING 2017 Department of Statistical Science Duke University FIRST YEAR EXAM - SPRING 017 Monday May 8th 017, 9:00 AM 1:00 PM NOTES: PLEASE READ CAREFULLY BEFORE BEGINNING EXAM! 1. Do not write solutions on the exam;

More information

Lecture 3. Truncation, length-bias and prevalence sampling

Lecture 3. Truncation, length-bias and prevalence sampling Lecture 3. Truncation, length-bias and prevalence sampling 3.1 Prevalent sampling Statistical techniques for truncated data have been integrated into survival analysis in last two decades. Truncation in

More information

LA PRISE DE CALAIS. çoys, çoys, har - dis. çoys, dis. tons, mantz, tons, Gas. c est. à ce. C est à ce. coup, c est à ce

LA PRISE DE CALAIS. çoys, çoys, har - dis. çoys, dis. tons, mantz, tons, Gas. c est. à ce. C est à ce. coup, c est à ce > ƒ? @ Z [ \ _ ' µ `. l 1 2 3 z Æ Ñ 6 = Ð l sl (~131 1606) rn % & +, l r s s, r 7 nr ss r r s s s, r s, r! " # $ s s ( ) r * s, / 0 s, r 4 r r 9;: < 10 r mnz, rz, r ns, 1 s ; j;k ns, q r s { } ~ l r mnz,

More information

TUTORIAL 8 SOLUTIONS #

TUTORIAL 8 SOLUTIONS # TUTORIAL 8 SOLUTIONS #9.11.21 Suppose that a single observation X is taken from a uniform density on [0,θ], and consider testing H 0 : θ = 1 versus H 1 : θ =2. (a) Find a test that has significance level

More information

EXCERPTS FROM ACTEX STUDY MANUAL FOR SOA EXAM P/CAS EXAM 1

EXCERPTS FROM ACTEX STUDY MANUAL FOR SOA EXAM P/CAS EXAM 1 EXCERPTS FROM ACTEX STUDY MANUAL FOR SOA EXAM P/CAS EXAM 1 Table of Contents Introductory Comments Section 2 - Conditional Probability and Independence TABLE OF CONTENTS INTRODUCTORY COMMENTS SECTION 0

More information

Dr. Maddah ENMG 617 EM Statistics 10/15/12. Nonparametric Statistics (2) (Goodness of fit tests)

Dr. Maddah ENMG 617 EM Statistics 10/15/12. Nonparametric Statistics (2) (Goodness of fit tests) Dr. Maddah ENMG 617 EM Statistics 10/15/12 Nonparametric Statistics (2) (Goodness of fit tests) Introduction Probability models used in decision making (Operations Research) and other fields require fitting

More information

Propagation of Uncertainty in a Maritime Risk Assessment Model utilizing Bayesian Simulation and Expert Judgment Aggregation Techniques

Propagation of Uncertainty in a Maritime Risk Assessment Model utilizing Bayesian Simulation and Expert Judgment Aggregation Techniques Presentation September 24 Rutgers University 1 Propagation of Uncertainty in a Maritime Risk Assessment Model utilizing Bayesian Simulation and Expert Judgment Aggregation Techniques V. Dinesh (VCU),T.A.

More information

HANDBOOK OF APPLICABLE MATHEMATICS

HANDBOOK OF APPLICABLE MATHEMATICS HANDBOOK OF APPLICABLE MATHEMATICS Chief Editor: Walter Ledermann Volume VI: Statistics PART A Edited by Emlyn Lloyd University of Lancaster A Wiley-Interscience Publication JOHN WILEY & SONS Chichester

More information

BAYESIAN ANALYSIS OF NONHOMOGENEOUS MARKOV CHAINS: APPLICATION TO MENTAL HEALTH DATA

BAYESIAN ANALYSIS OF NONHOMOGENEOUS MARKOV CHAINS: APPLICATION TO MENTAL HEALTH DATA BAYESIAN ANALYSIS OF NONHOMOGENEOUS MARKOV CHAINS: APPLICATION TO MENTAL HEALTH DATA " # $ Minje Sung, Refik Soyer, Nguyen Nhan " Department of Biostatistics and Bioinformatics, Box 3454, Duke University

More information

Two-stage Adaptive Randomization for Delayed Response in Clinical Trials

Two-stage Adaptive Randomization for Delayed Response in Clinical Trials Two-stage Adaptive Randomization for Delayed Response in Clinical Trials Guosheng Yin Department of Statistics and Actuarial Science The University of Hong Kong Joint work with J. Xu PSI and RSS Journal

More information

11 Survival Analysis and Empirical Likelihood

11 Survival Analysis and Empirical Likelihood 11 Survival Analysis and Empirical Likelihood The first paper of empirical likelihood is actually about confidence intervals with the Kaplan-Meier estimator (Thomas and Grunkmeier 1979), i.e. deals with

More information

Modelling the risk process

Modelling the risk process Modelling the risk process Krzysztof Burnecki Hugo Steinhaus Center Wroc law University of Technology www.im.pwr.wroc.pl/ hugo Modelling the risk process 1 Risk process If (Ω, F, P) is a probability space

More information

Lund Institute of Technology Centre for Mathematical Sciences Mathematical Statistics

Lund Institute of Technology Centre for Mathematical Sciences Mathematical Statistics Lund Institute of Technology Centre for Mathematical Sciences Mathematical Statistics STATISTICAL METHODS FOR SAFETY ANALYSIS FMS065 ÓÑÔÙØ Ö Ü Ö Ì ÓÓØ ØÖ Ô Ð ÓÖ Ø Ñ Ò Ý Ò Ò ÐÝ In this exercise we will

More information

Survival Analysis Math 434 Fall 2011

Survival Analysis Math 434 Fall 2011 Survival Analysis Math 434 Fall 2011 Part IV: Chap. 8,9.2,9.3,11: Semiparametric Proportional Hazards Regression Jimin Ding Math Dept. www.math.wustl.edu/ jmding/math434/fall09/index.html Basic Model Setup

More information

Recall the Basics of Hypothesis Testing

Recall the Basics of Hypothesis Testing Recall the Basics of Hypothesis Testing The level of significance α, (size of test) is defined as the probability of X falling in w (rejecting H 0 ) when H 0 is true: P(X w H 0 ) = α. H 0 TRUE H 1 TRUE

More information

COPYRIGHTED MATERIAL CONTENTS. Preface Preface to the First Edition

COPYRIGHTED MATERIAL CONTENTS. Preface Preface to the First Edition Preface Preface to the First Edition xi xiii 1 Basic Probability Theory 1 1.1 Introduction 1 1.2 Sample Spaces and Events 3 1.3 The Axioms of Probability 7 1.4 Finite Sample Spaces and Combinatorics 15

More information

Digital Communications: An Overview of Fundamentals

Digital Communications: An Overview of Fundamentals IMPERIAL COLLEGE of SCIENCE, TECHNOLOGY and MEDICINE, DEPARTMENT of ELECTRICAL and ELECTRONIC ENGINEERING. COMPACT LECTURE NOTES on COMMUNICATION THEORY. Prof Athanassios Manikas, Spring 2001 Digital Communications:

More information

Review. DS GA 1002 Statistical and Mathematical Models. Carlos Fernandez-Granda

Review. DS GA 1002 Statistical and Mathematical Models.   Carlos Fernandez-Granda Review DS GA 1002 Statistical and Mathematical Models http://www.cims.nyu.edu/~cfgranda/pages/dsga1002_fall16 Carlos Fernandez-Granda Probability and statistics Probability: Framework for dealing with

More information

Systems of Equations 1. Systems of Linear Equations

Systems of Equations 1. Systems of Linear Equations Lecture 1 Systems of Equations 1. Systems of Linear Equations [We will see examples of how linear equations arise here, and how they are solved:] Example 1: In a lab experiment, a researcher wants to provide

More information

: œ Ö: =? À =ß> real numbers. œ the previous plane with each point translated by : Ðfor example,! is translated to :)

: œ Ö: =? À =ß> real numbers. œ the previous plane with each point translated by : Ðfor example,! is translated to :) â SpanÖ?ß@ œ Ö =? > @ À =ß> real numbers : SpanÖ?ß@ œ Ö: =? > @ À =ß> real numbers œ the previous plane with each point translated by : Ðfor example, is translated to :) á In general: Adding a vector :

More information

STATS 200: Introduction to Statistical Inference. Lecture 29: Course review

STATS 200: Introduction to Statistical Inference. Lecture 29: Course review STATS 200: Introduction to Statistical Inference Lecture 29: Course review Course review We started in Lecture 1 with a fundamental assumption: Data is a realization of a random process. The goal throughout

More information

Method of Moments. which we usually denote by X or sometimes by X n to emphasize that there are n observations.

Method of Moments. which we usually denote by X or sometimes by X n to emphasize that there are n observations. Method of Moments Definition. If {X 1,..., X n } is a sample from a population, then the empirical k-th moment of this sample is defined to be X k 1 + + Xk n n Example. For a sample {X 1, X, X 3 } the

More information

Statistical Inference and Methods

Statistical Inference and Methods Department of Mathematics Imperial College London d.stephens@imperial.ac.uk http://stats.ma.ic.ac.uk/ das01/ 31st January 2006 Part VI Session 6: Filtering and Time to Event Data Session 6: Filtering and

More information

1 Glivenko-Cantelli type theorems

1 Glivenko-Cantelli type theorems STA79 Lecture Spring Semester Glivenko-Cantelli type theorems Given i.i.d. observations X,..., X n with unknown distribution function F (t, consider the empirical (sample CDF ˆF n (t = I [Xi t]. n Then

More information

Parameter Estimation for Partially Complete Time and Type of Failure Data

Parameter Estimation for Partially Complete Time and Type of Failure Data Biometrical Journal 46 (004), 65 79 DOI 0.00/bimj.00004 arameter Estimation for artially Complete Time and Type of Failure Data Debasis Kundu Department of Mathematics, Indian Institute of Technology Kanpur,

More information

Cellular Automaton Growth on # : Theorems, Examples, and Problems

Cellular Automaton Growth on # : Theorems, Examples, and Problems Cellular Automaton Growth on : Theorems, Examples, and Problems (Excerpt from Advances in Applied Mathematics) Exactly 1 Solidification We will study the evolution starting from a single occupied cell

More information

probability of k samples out of J fall in R.

probability of k samples out of J fall in R. Nonparametric Techniques for Density Estimation (DHS Ch. 4) n Introduction n Estimation Procedure n Parzen Window Estimation n Parzen Window Example n K n -Nearest Neighbor Estimation Introduction Suppose

More information