Introduction to Rare Event Simulation

Size: px
Start display at page:

Download "Introduction to Rare Event Simulation"

Transcription

1 Introduction to Rare Event Simulation Brown University: Summer School on Rare Event Simulation Jose Blanchet Columbia University. Department of Statistics, Department of IEOR. Blanchet (Columbia) 1 / 31

2 Goals / questions to address in this lecture The need for rare event simulation by introducing several problems. Blanchet (Columbia) 2 / 31

3 Goals / questions to address in this lecture The need for rare event simulation by introducing several problems. Why do we care about rare event simulation? Blanchet (Columbia) 2 / 31

4 Goals / questions to address in this lecture The need for rare event simulation by introducing several problems. Why do we care about rare event simulation? How does one measure the effi ciency of rare event simulation algorithms? Blanchet (Columbia) 2 / 31

5 Goals / questions to address in this lecture The need for rare event simulation by introducing several problems. Why do we care about rare event simulation? How does one measure the effi ciency of rare event simulation algorithms? Describe basic rare event simulation techniques. Blanchet (Columbia) 2 / 31

6 Why are Rare Events Diffi cult to Assess? Typically no closed forms But crudely implemented simulation might not be good Blanchet (Columbia) 3 / 31

7 Why are Rare Events Diffi cult to Assess? Blanchet (Columbia) 4 / 31

8 Why are Rare Events Diffi cult to Assess? Relative mean squared error (MSE) PER TRIAL = stdev / mean P (RED) (1 P (RED)) P (RED) 1 P (RED). Blanchet (Columbia) 5 / 31

9 Why are Rare Events Diffi cult to Assess? Relative mean squared error (MSE) PER TRIAL = stdev / mean P (RED) (1 P (RED)) P (RED) 1 P (RED). Moral of Story: Naive Monte Carlo needs ( ) 1 N = O P (Rare Event) replications to obtain a given relative accuracy. Blanchet (Columbia) 5 / 31

10 Introduction to Rare Event Simulation Rare event simulation = techniques for effi cient estimation of rare-event probabilities AND conditional expectations given such rare events. Blanchet (Columbia) 6 / 31

11 Introduction to Rare Event Simulation Rare event simulation = techniques for effi cient estimation of rare-event probabilities AND conditional expectations given such rare events. Arises in settings such as: Blanchet (Columbia) 6 / 31

12 Introduction to Rare Event Simulation Rare event simulation = techniques for effi cient estimation of rare-event probabilities AND conditional expectations given such rare events. Arises in settings such as: 1 Finance and Insurance, Blanchet (Columbia) 6 / 31

13 Introduction to Rare Event Simulation Rare event simulation = techniques for effi cient estimation of rare-event probabilities AND conditional expectations given such rare events. Arises in settings such as: 1 Finance and Insurance, 2 Queueing networks (operations and communications), Blanchet (Columbia) 6 / 31

14 Introduction to Rare Event Simulation Rare event simulation = techniques for effi cient estimation of rare-event probabilities AND conditional expectations given such rare events. Arises in settings such as: 1 Finance and Insurance, 2 Queueing networks (operations and communications), 3 Environmental and physical sciences, Blanchet (Columbia) 6 / 31

15 Introduction to Rare Event Simulation Rare event simulation = techniques for effi cient estimation of rare-event probabilities AND conditional expectations given such rare events. Arises in settings such as: 1 Finance and Insurance, 2 Queueing networks (operations and communications), 3 Environmental and physical sciences, 4 Statistical inference and combinatorics... Blanchet (Columbia) 6 / 31

16 Introduction to Rare Event Simulation Rare event simulation = techniques for effi cient estimation of rare-event probabilities AND conditional expectations given such rare events. Arises in settings such as: 1 Finance and Insurance, 2 Queueing networks (operations and communications), 3 Environmental and physical sciences, 4 Statistical inference and combinatorics... 5 Stochastic optimization, Blanchet (Columbia) 6 / 31

17 Insurance Reserve Setup V i s i.i.d. (independent and identially distributed) claim sizes Blanchet (Columbia) 7 / 31

18 Insurance Reserve Setup V i s i.i.d. (independent and identially distributed) claim sizes τ i s i.i.d. inter-arrival times Blanchet (Columbia) 7 / 31

19 Insurance Reserve Setup V i s i.i.d. (independent and identially distributed) claim sizes τ i s i.i.d. inter-arrival times A n = time of the n-th arrival Blanchet (Columbia) 7 / 31

20 Insurance Reserve Setup V i s i.i.d. (independent and identially distributed) claim sizes τ i s i.i.d. inter-arrival times A n = time of the n-th arrival Constant premium p Blanchet (Columbia) 7 / 31

21 Insurance Reserve Setup V i s i.i.d. (independent and identially distributed) claim sizes τ i s i.i.d. inter-arrival times A n = time of the n-th arrival Constant premium p Reserve process N (t) R (t) = b + pt j=1 V j Blanchet (Columbia) 7 / 31

22 Insurance Reserve Setup V i s i.i.d. (independent and identially distributed) claim sizes τ i s i.i.d. inter-arrival times A n = time of the n-th arrival Constant premium p Reserve process N (t) = # arrivals up to time t N (t) R (t) = b + pt j=1 V j Blanchet (Columbia) 7 / 31

23 Insurance Reserve Plot Plot of risk reserve R(t) b A 1 A 2 A 3 t Blanchet (Columbia) 8 / 31

24 Insurance Reserve and Random Walks Evaluating reserve at arrival times we get random walk Blanchet (Columbia) 9 / 31

25 Insurance Reserve and Random Walks Evaluating reserve at arrival times we get random walk Suppose Y 1, Y 2,... are i.i.d. S (n) = b + Y Y n Blanchet (Columbia) 9 / 31

26 Insurance Reserve and Random Walks Evaluating reserve at arrival times we get random walk Suppose Y 1, Y 2,... are i.i.d. S (n) = b + Y Y n R (A n ) = S (n) reserve at arrival times with Y n = pτ n V n. R(t) b A 1 A 2 A 3 t Blanchet (Columbia) 9 / 31

27 Insurance Reserve and Random Walks Evaluating reserve at arrival times we get random walk Suppose Y 1, Y 2,... are i.i.d. S (n) = b + Y Y n R (A n ) = S (n) reserve at arrival times with Y n = pτ n V n. R(t) b A 1 A 2 A 3 t Question: What is the chance that inf{r (t) : t 0} < 0? And how does this happen? Blanchet (Columbia) 9 / 31

28 How Does Ruin Occur with Light-Tailed Increments? Y i s are N(0, 1), EY i = 1 Blanchet (Columbia) 10 / 31

29 How Does Ruin Occur with Light-Tailed Increments? Y i s are N(0, 1), EY i = 1 Random walk conditioned on ruin Blanchet (Columbia) 10 / 31

30 How Does Ruin Occur with Light-Tailed Increments? Y i s are N(0, 1), EY i = 1 Random walk conditioned on ruin Light tails: Exponential, Gamma, Gaussian, mixtures of these, etc. Blanchet (Columbia) 10 / 31

31 How Does Ruin Occur with Light-Tailed Increments? Y i s are N(0, 1), EY i = 1 Random walk conditioned on ruin Light tails: Exponential, Gamma, Gaussian, mixtures of these, etc. Picture generated with Siegmund s 76 algorithm Blanchet (Columbia) 10 / 31

32 Back to Heavy-tailed Insurance Ruin Ruin with with Power law Pareto Type Claims Increments 400 Random Reserve Walk Value B Time Y i s are t-distributed, EY i = 1, Var (Y i ) = 1, 4 degrees of freedom. Blanchet (Columbia) 11 / 31

33 Back to Heavy-tailed Insurance Ruin Ruin with with Power law Pareto Type Claims Increments 400 Random Reserve Walk Value B Time Y i s are t-distributed, EY i = 1, Var (Y i ) = 1, 4 degrees of freedom. Heavy-tailed random walk conditioned on ruin. Blanchet (Columbia) 11 / 31

34 Back to Heavy-tailed Insurance Ruin Ruin with with Power law Pareto Type Claims Increments 400 Random Reserve Walk Value B Time Y i s are t-distributed, EY i = 1, Var (Y i ) = 1, 4 degrees of freedom. Heavy-tailed random walk conditioned on ruin. Picture generated with Blanchet and Glynn s 09 algorithm. Blanchet (Columbia) 11 / 31

35 A Two-node Tandem Network Poisson arrivals, exponential service times, and Markovian routing... Blanchet (Columbia) 12 / 31

36 Example: A Simple Stochastic Network Questions of interest: How would the system perform IF TOTAL POPULATION reaches n inside a busy period? How likely is this event under different parameters / designs? Blanchet (Columbia) 13 / 31

37 Naive Benchmark: Crude Monte Carlo Say λ =.1, µ 1 =.5, µ 2 =.4, p =.1 Blanchet (Columbia) 14 / 31

38 Naive Benchmark: Crude Monte Carlo Say λ =.1, µ 1 =.5, µ 2 =.4, p =.1 Overflow probability 10 7 Blanchet (Columbia) 14 / 31

39 Naive Benchmark: Crude Monte Carlo Say λ =.1, µ 1 =.5, µ 2 =.4, p =.1 Overflow probability 10 7 Each replication in crude Monte Carlo 10 3 secs Blanchet (Columbia) 14 / 31

40 Naive Benchmark: Crude Monte Carlo Say λ =.1, µ 1 =.5, µ 2 =.4, p =.1 Overflow probability 10 7 Each replication in crude Monte Carlo 10 3 secs Time to estimate with 10% precision 10 7 /1000 = 10 4 secs 2.7 hours Blanchet (Columbia) 14 / 31

41 Naive Benchmark: Crude Monte Carlo Say λ =.1, µ 1 =.5, µ 2 =.4, p =.1 Overflow probability 10 7 Each replication in crude Monte Carlo 10 3 secs Time to estimate with 10% precision 10 7 /1000 = 10 4 secs 2.7 hours What if you want to do sensitivity analysis for a wide range of parameters? Blanchet (Columbia) 14 / 31

42 Naive Benchmark: Crude Monte Carlo Say λ =.1, µ 1 =.5, µ 2 =.4, p =.1 Overflow probability 10 7 Each replication in crude Monte Carlo 10 3 secs Time to estimate with 10% precision 10 7 /1000 = 10 4 secs 2.7 hours What if you want to do sensitivity analysis for a wide range of parameters? What about different network designs? Blanchet (Columbia) 14 / 31

43 Rare Events in Networks Question: How does the TOTAL CONTENT OF THE SYSTEM is most likely to reach a give level in an operation cycle? Blanchet (Columbia) 15 / 31

44 Rare Events in Networks Question: How does the TOTAL CONTENT OF THE SYSTEM is most likely to reach a give level in an operation cycle? Say λ =.2, µ 1 =.3, µ 2 =.5... How can total content reach 50? Blanchet (Columbia) 15 / 31

45 Rare Events in Networks Question: How does the TOTAL CONTENT OF THE SYSTEM is most likely to reach a give level in an operation cycle? Say λ =.2, µ 1 =.3, µ 2 =.5... How can total content reach 50? Naive Monte Carlo takes 115 days Blanchet (Columbia) 15 / 31

46 Rare Events in Networks Question: How does the TOTAL CONTENT OF THE SYSTEM is most likely to reach a give level in an operation cycle? Say λ =.2, µ 1 =.3, µ 2 =.5... How can total content reach 50? Naive Monte Carlo takes 115 days Each picture below took.01 seconds Blanchet (Columbia) 15 / 31

47 Rare Events in Networks λ =.1, µ 1 =.3, µ 2 =.5 Plot of Conditional Path X 2 OF X 1 Blanchet (Columbia) 16 / 31

48 Rare Events in Networks λ =.1, µ 1 =.5, µ 2 =.2 Plot of Conditional Path OF X 2 X 1 Blanchet (Columbia) 17 / 31

49 Rare Events in Networks λ =.1, µ 1 =.4, µ 2 =.4 Plot of Conditional Path X 2 OF X 1 Blanchet (Columbia) 18 / 31

50 Several Notions of Effi ciency Assume the estimator is unbiased. Blanchet (Columbia) 19 / 31

51 Several Notions of Effi ciency Assume the estimator is unbiased. E (Z ) = P (A) and we assume P (A) 0. Blanchet (Columbia) 19 / 31

52 Several Notions of Effi ciency Assume the estimator is unbiased. E (Z ) = P (A) and we assume P (A) 0. Logarithmic Effi ciency (weakly effi ciency or asymptotic optimality): log E ( Z 2) lim P (A) 0 2 log P (A) = 1. Blanchet (Columbia) 19 / 31

53 Several Notions of Effi ciency Assume the estimator is unbiased. E (Z ) = P (A) and we assume P (A) 0. Logarithmic Effi ciency (weakly effi ciency or asymptotic optimality): log E ( Z 2) lim P (A) 0 2 log P (A) = 1. Bounded Relative Error E ( Z 2) = O 2 (1) as P (A) 0. P (A) Blanchet (Columbia) 19 / 31

54 Effi ciency of Naive Monte Carlo If Z = I (A) then E ( Z 2) = E (Z ) = P (A) Blanchet (Columbia) 20 / 31

55 Effi ciency of Naive Monte Carlo If Z = I (A) then E ( Z 2) = E (Z ) = P (A) Therefore lim P (A) 0 log P (A) 2 log P (A) = 1 2. Blanchet (Columbia) 20 / 31

56 Effi ciency of Naive Monte Carlo If Z = I (A) then E ( Z 2) = E (Z ) = P (A) Therefore lim P (A) 0 log P (A) 2 log P (A) = 1 2. How to design effi cient rare event simulation estimators? Importance Sampling and other variance reduction techniques... Blanchet (Columbia) 20 / 31

57 Graphical Interpretation of Importance Sampling Importance sampling (I.S.): sample from the important region and correct via likelihood ratio RED AREA PROPORTION DARTS IN RED AREA Blanchet (Columbia) 21 / 31

58 Importance Sampling Goal: Estimate P (A) > 0 Blanchet (Columbia) 22 / 31

59 Importance Sampling Goal: Estimate P (A) > 0 Choose P and simulate ω from it Blanchet (Columbia) 22 / 31

60 Importance Sampling Goal: Estimate P (A) > 0 Choose P and simulate ω from it Importance Sampling (I.S.) estimator per trial is I.S.Estimator = L (ω) I (ω A), where L (ω) is the likelihood ratio (i.e. L (ω) = dp (ω) /d P (ω)). Blanchet (Columbia) 22 / 31

61 Importance Sampling Goal: Estimate P (A) > 0 Choose P and simulate ω from it Importance Sampling (I.S.) estimator per trial is I.S.Estimator = L (ω) I (ω A), where L (ω) is the likelihood ratio (i.e. L (ω) = dp (ω) /d P (ω)). NOTE: P ( ) is called a change-of-measure Blanchet (Columbia) 22 / 31

62 Importance Sampling Suppose we choose P ( ) = P ( A) L (ω) = dp (ω) I (ω A) = P (A) I (ω A) dp (ω) /P (A) Blanchet (Columbia) 23 / 31

63 Importance Sampling Suppose we choose P ( ) = P ( A) L (ω) = dp (ω) I (ω A) = P (A) I (ω A) dp (ω) /P (A) Estimator has zero variance, but requires knoweledge of P (A) Blanchet (Columbia) 23 / 31

64 Importance Sampling Suppose we choose P ( ) = P ( A) L (ω) = dp (ω) I (ω A) = P (A) I (ω A) dp (ω) /P (A) Estimator has zero variance, but requires knoweledge of P (A) Lesson: Try choosing P ( ) close to P ( A)! Blanchet (Columbia) 23 / 31

65 Large Deviations of the Empirical Mean Suppose X 1, X 2,... are i.i.d. and H (θ) = log E exp (θx ) < for θ in a neighborhood of the origin. Blanchet (Columbia) 24 / 31

66 Large Deviations of the Empirical Mean Suppose X 1, X 2,... are i.i.d. and H (θ) = log E exp (θx ) < for θ in a neighborhood of the origin. Assume that E (X i ) = 0. Blanchet (Columbia) 24 / 31

67 Large Deviations of the Empirical Mean Suppose X 1, X 2,... are i.i.d. and H (θ) = log E exp (θx ) < for θ in a neighborhood of the origin. Assume that E (X i ) = 0. Suppose that for each a > 0, there is θ a such that H (θ a ) = a. Blanchet (Columbia) 24 / 31

68 Large Deviations of the Empirical Mean Suppose X 1, X 2,... are i.i.d. and H (θ) = log E exp (θx ) < for θ in a neighborhood of the origin. Assume that E (X i ) = 0. Suppose that for each a > 0, there is θ a such that H (θ a ) = a. Consider the problem of estimating via simulation where S (n) = X X n. P (S (n) > na), Blanchet (Columbia) 24 / 31

69 Exponential Tilting Suppose that X i has density f ( ) and define f θ (x) = exp (θx H (θ)) f (x). Blanchet (Columbia) 25 / 31

70 Exponential Tilting Suppose that X i has density f ( ) and define f θ (x) = exp (θx H (θ)) f (x). Note that log E θ (exp (ηx )) = log f θ (x) exp (ηx) dx = log f (x) exp ((η + θ) x H (θ)) dx = H (η + θ) H (θ). Blanchet (Columbia) 25 / 31

71 Exponential Tilting Suppose that X i has density f ( ) and define f θ (x) = exp (θx H (θ)) f (x). Note that log E θ (exp (ηx )) = log f θ (x) exp (ηx) dx = log f (x) exp ((η + θ) x H (θ)) dx = H (η + θ) H (θ). Therefore E θ (X ) = H (θ) and Var θ (X ) = H (θ). Blanchet (Columbia) 25 / 31

72 Exponential Tilting in Action Cramer s theorem (strong version): (Sketch) = = P (S n > na) f (x 1 )...f (x n ) I (s n > na) dx 1:n f θa (x 1 )...f θa (x n ) exp ( θ a s n + nh (θ a )) dx 1:n {s n >na} = exp ( n[aθ a H (θ a )]) E θa [exp ( θ a (S n na)) I (S n > na)] = exp ( ni (a)) E θa [exp ( θ a (S n na)) I (S n > na)]. Blanchet (Columbia) 26 / 31

73 Exponential Tilting in Action Cramer s theorem (strong version): (Sketch) = = P (S n > na) f (x 1 )...f (x n ) I (s n > na) dx 1:n f θa (x 1 )...f θa (x n ) exp ( θ a s n + nh (θ a )) dx 1:n {s n >na} = exp ( n[aθ a H (θ a )]) E θa [exp ( θ a (S n na)) I (S n > na)] = exp ( ni (a)) E θa [exp ( θ a (S n na)) I (S n > na)]. Use CLT and approximate (S n na) by N (0, nh (θ a )) to obtain E θa [exp ( θ a (S n na)) I (S n > na)] 1 n 1/2 (2πH (θ a )) 1/2 θ a. Blanchet (Columbia) 26 / 31

74 Why is Exponential tilting Useful? Theorem Under the assumptions imposed Proof. P (X 1 dx 1 S n > na) P θa (X 1 dx 1 ). In simple words, exponential tilting approximates the zero-variance (optimal!) change-of-measure! P (X 1 dx 1 S n > na) = P (S n 1 > na x 1 ) f (x 1 ) dx. P (S n > na) Now apply Cramer s theorem with a (na x 1 ) /(n 1) and n n 1 in the numerator and simplify, observing that ( ) ( na x1 θ = θ a + a x ) 1 1 a x 1 θ (a) + n 1 n 1 H (θ a ) n 1. Blanchet (Columbia) 27 / 31

75 First Asymptotically Optimal Rare Event Simulation Estimator Recall the identity used in Cramer s theorem P (S n > na) = f (x 1 )...f (x n ) I (s n > na) dx 1:n = f θa (x 1 )...f θa (x n ) exp ( θ a s n + nh (θ a )) dx 1:n {s n >na} = E θa [exp ( θ a S n + nh (θ a )) I (S n > na)]. Blanchet (Columbia) 28 / 31

76 First Asymptotically Optimal Rare Event Simulation Estimator Recall the identity used in Cramer s theorem P (S n > na) = f (x 1 )...f (x n ) I (s n > na) dx 1:n = f θa (x 1 )...f θa (x n ) exp ( θ a s n + nh (θ a )) dx 1:n {s n >na} = E θa [exp ( θ a S n + nh (θ a )) I (S n > na)]. IMPORTANCE SAMPLING ESTIMATOR: Z = exp ( θ a S n + nh (θ a )) I (S n > na), with X 1,..., X n simulated using the density f θa ( ). Blanchet (Columbia) 28 / 31

77 First Asymptotically Optimal Rare Event Simulation Estimator Recall the identity used in Cramer s theorem P (S n > na) = f (x 1 )...f (x n ) I (s n > na) dx 1:n = f θa (x 1 )...f θa (x n ) exp ( θ a s n + nh (θ a )) dx 1:n {s n >na} = E θa [exp ( θ a S n + nh (θ a )) I (S n > na)]. IMPORTANCE SAMPLING ESTIMATOR: Z = exp ( θ a S n + nh (θ a )) I (S n > na), with X 1,..., X n simulated using the density f θa ( ). Let us verify now that Z is asymptotically effi cient! Blanchet (Columbia) 28 / 31

78 First Asymptotically Optimal Rare Event Simulation Estimator IMPORTANCE SAMPLING ESTIMATOR: Z = exp ( θ a S n + nh (θ a )) I (S n > na), with X 1,..., X n simulated using the density f θa ( ). Blanchet (Columbia) 29 / 31

79 First Asymptotically Optimal Rare Event Simulation Estimator IMPORTANCE SAMPLING ESTIMATOR: Z = exp ( θ a S n + nh (θ a )) I (S n > na), with X 1,..., X n simulated using the density f θa ( ). Therefore, ( E θa Z 2 ) = E θa exp ( 2θ a S n + 2nH (θ a )) I (S n > na) = exp ( 2nI (a)) E θa [exp ( 2θ a (S n na)) I (S n > na)]. Blanchet (Columbia) 29 / 31

80 First Asymptotically Optimal Rare Event Simulation Estimator IMPORTANCE SAMPLING ESTIMATOR: Z = exp ( θ a S n + nh (θ a )) I (S n > na), with X 1,..., X n simulated using the density f θa ( ). Therefore, ( E θa Z 2 ) = E θa exp ( 2θ a S n + 2nH (θ a )) I (S n > na) = exp ( 2nI (a)) E θa [exp ( 2θ a (S n na)) I (S n > na)]. Thus ( log E θa Z 2 ) 2nI (a) = + o (1) 1. 2 log P (S n > na) 2nI (a) Blanchet (Columbia) 29 / 31

81 How to Simulate under Exponential Tilting Key is to identify: H θ (η) = H (η + θ) H (θ). Blanchet (Columbia) 30 / 31

82 How to Simulate under Exponential Tilting Key is to identify: H θ (η) = H (η + θ) H (θ). Example 1: Gaussian if X N (µ, 1) H (θ) = θ 2 /2 + µθ, H θ (η) = η 2 /2 + (θ + µ) η. Thus, under f θ, X N (µ, 1). Blanchet (Columbia) 30 / 31

83 How to Simulate under Exponential Tilting Key is to identify: H θ (η) = H (η + θ) H (θ). Example 1: Gaussian if X N (µ, 1) H (θ) = θ 2 /2 + µθ, H θ (η) = η 2 /2 + (θ + µ) η. Thus, under f θ, X N (µ, 1). Example 2: If X Poisson (λ), then H (θ) = λ (exp (θ) 1), H θ (η) = λ exp (θ) (exp (η) 1). Thus, under f θ, X Poisson (λ exp (θ)). Blanchet (Columbia) 30 / 31

84 Conclusions 1 Rare event simulation is a power numerical tool applicable in a wide range of examples. Blanchet (Columbia) 31 / 31

85 Conclusions 1 Rare event simulation is a power numerical tool applicable in a wide range of examples. 2 Importance sampling can be used to construct effi cient estimators. Blanchet (Columbia) 31 / 31

86 Conclusions 1 Rare event simulation is a power numerical tool applicable in a wide range of examples. 2 Importance sampling can be used to construct effi cient estimators. 3 Best importance sampler is the conditional distribution given the event of interest. Blanchet (Columbia) 31 / 31

87 Conclusions 1 Rare event simulation is a power numerical tool applicable in a wide range of examples. 2 Importance sampling can be used to construct effi cient estimators. 3 Best importance sampler is the conditional distribution given the event of interest. 4 Effi cient algorithms come basically in two flavors: logarithmically and strongly effi cient. Blanchet (Columbia) 31 / 31

88 Conclusions 1 Rare event simulation is a power numerical tool applicable in a wide range of examples. 2 Importance sampling can be used to construct effi cient estimators. 3 Best importance sampler is the conditional distribution given the event of interest. 4 Effi cient algorithms come basically in two flavors: logarithmically and strongly effi cient. 5 Exponential tilting can be used to approximate the conditional distribution given the event of interest. Blanchet (Columbia) 31 / 31

E cient Monte Carlo for Gaussian Fields and Processes

E cient Monte Carlo for Gaussian Fields and Processes E cient Monte Carlo for Gaussian Fields and Processes Jose Blanchet (with R. Adler, J. C. Liu, and C. Li) Columbia University Nov, 2010 Jose Blanchet (Columbia) Monte Carlo for Gaussian Fields Nov, 2010

More information

Rare-Event Simulation

Rare-Event Simulation Rare-Event Simulation Background: Read Chapter 6 of text. 1 Why is Rare-Event Simulation Challenging? Consider the problem of computing α = P(A) when P(A) is small (i.e. rare ). The crude Monte Carlo estimator

More information

Importance Sampling Methodology for Multidimensional Heavy-tailed Random Walks

Importance Sampling Methodology for Multidimensional Heavy-tailed Random Walks Importance Sampling Methodology for Multidimensional Heavy-tailed Random Walks Jose Blanchet (joint work with Jingchen Liu) Columbia IEOR Department RESIM Blanchet (Columbia) IS for Heavy-tailed Walks

More information

State-dependent Importance Sampling for Rare-event Simulation: An Overview and Recent Advances

State-dependent Importance Sampling for Rare-event Simulation: An Overview and Recent Advances State-dependent Importance Sampling for Rare-event Simulation: An Overview and Recent Advances By Jose Blanchet and Henry Lam Columbia University and Boston University February 7, 2011 Abstract This paper

More information

Large deviations for random walks under subexponentiality: the big-jump domain

Large deviations for random walks under subexponentiality: the big-jump domain Large deviations under subexponentiality p. Large deviations for random walks under subexponentiality: the big-jump domain Ton Dieker, IBM Watson Research Center joint work with D. Denisov (Heriot-Watt,

More information

Efficient rare-event simulation for sums of dependent random varia

Efficient rare-event simulation for sums of dependent random varia Efficient rare-event simulation for sums of dependent random variables Leonardo Rojas-Nandayapa joint work with José Blanchet February 13, 2012 MCQMC UNSW, Sydney, Australia Contents Introduction 1 Introduction

More information

On Lyapunov Inequalities and Subsolutions for Efficient Importance Sampling

On Lyapunov Inequalities and Subsolutions for Efficient Importance Sampling On Lyapunov Inequalities and Subsolutions for Efficient Importance Sampling By Jose Blanchet, Kevin Leder and Peter Glynn Columbia University, Columbia University and Stanford University July 28, 2009

More information

EFFICIENT TAIL ESTIMATION FOR SUMS OF CORRELATED LOGNORMALS. Leonardo Rojas-Nandayapa Department of Mathematical Sciences

EFFICIENT TAIL ESTIMATION FOR SUMS OF CORRELATED LOGNORMALS. Leonardo Rojas-Nandayapa Department of Mathematical Sciences Proceedings of the 2008 Winter Simulation Conference S. J. Mason, R. R. Hill, L. Mönch, O. Rose, T. Jefferson, J. W. Fowler eds. EFFICIENT TAIL ESTIMATION FOR SUMS OF CORRELATED LOGNORMALS Jose Blanchet

More information

ABSTRACT 1 INTRODUCTION. Jose Blanchet. Henry Lam. Boston University 111 Cummington Street Boston, MA 02215, USA

ABSTRACT 1 INTRODUCTION. Jose Blanchet. Henry Lam. Boston University 111 Cummington Street Boston, MA 02215, USA Proceedings of the 2011 Winter Simulation Conference S. Jain, R. R. Creasey, J. Himmelspach, K. P. White, and M. Fu, eds. RARE EVENT SIMULATION TECHNIQUES Jose Blanchet Columbia University 500 W 120th

More information

Efficient Rare-event Simulation for Perpetuities

Efficient Rare-event Simulation for Perpetuities Efficient Rare-event Simulation for Perpetuities Blanchet, J., Lam, H., and Zwart, B. We consider perpetuities of the form Abstract D = B 1 exp Y 1 ) + B 2 exp Y 1 + Y 2 ) +..., where the Y j s and B j

More information

On the inefficiency of state-independent importance sampling in the presence of heavy tails

On the inefficiency of state-independent importance sampling in the presence of heavy tails Operations Research Letters 35 (2007) 251 260 Operations Research Letters www.elsevier.com/locate/orl On the inefficiency of state-independent importance sampling in the presence of heavy tails Achal Bassamboo

More information

Rare event simulation for the ruin problem with investments via importance sampling and duality

Rare event simulation for the ruin problem with investments via importance sampling and duality Rare event simulation for the ruin problem with investments via importance sampling and duality Jerey Collamore University of Copenhagen Joint work with Anand Vidyashankar (GMU) and Guoqing Diao (GMU).

More information

Rare-event Simulation Techniques: An Introduction and Recent Advances

Rare-event Simulation Techniques: An Introduction and Recent Advances Rare-event Simulation Techniques: An Introduction and Recent Advances S. Juneja Tata Institute of Fundamental Research, India juneja@tifr.res.in P. Shahabuddin Columbia University perwez@ieor.columbia.edu

More information

Variance Reduction s Greatest Hits

Variance Reduction s Greatest Hits 1 Variance Reduction s Greatest Hits Pierre L Ecuyer CIRRELT, GERAD, and Département d Informatique et de Recherche Opérationnelle Université de Montréal, Canada ESM 2007, Malta, October 22, 2007 Context

More information

Proceedings of the 2011 Winter Simulation Conference S. Jain, R. R. Creasey, J. Himmelspach, K. P. White, and M. Fu, eds.

Proceedings of the 2011 Winter Simulation Conference S. Jain, R. R. Creasey, J. Himmelspach, K. P. White, and M. Fu, eds. Proceedings of the 2011 Winter Simulation Conference S. Jain, R. R. Creasey, J. Himmelspach, K. P. White, and M. Fu, eds. IMPORTANCE SAMPLING FOR ACTUARIAL COST ANALYSIS UNDER A HEAVY TRAFFIC MODEL Jose

More information

Standard Error of Technical Cost Incorporating Parameter Uncertainty

Standard Error of Technical Cost Incorporating Parameter Uncertainty Standard Error of Technical Cost Incorporating Parameter Uncertainty Christopher Morton Insurance Australia Group This presentation has been prepared for the Actuaries Institute 2012 General Insurance

More information

Asymptotics and Simulation of Heavy-Tailed Processes

Asymptotics and Simulation of Heavy-Tailed Processes Asymptotics and Simulation of Heavy-Tailed Processes Department of Mathematics Stockholm, Sweden Workshop on Heavy-tailed Distributions and Extreme Value Theory ISI Kolkata January 14-17, 2013 Outline

More information

IEOR 6711: Stochastic Models I SOLUTIONS to the First Midterm Exam, October 7, 2008

IEOR 6711: Stochastic Models I SOLUTIONS to the First Midterm Exam, October 7, 2008 IEOR 6711: Stochastic Models I SOLUTIONS to the First Midterm Exam, October 7, 2008 Justify your answers; show your work. 1. A sequence of Events. (10 points) Let {B n : n 1} be a sequence of events in

More information

Fluid Heuristics, Lyapunov Bounds and E cient Importance Sampling for a Heavy-tailed G/G/1 Queue

Fluid Heuristics, Lyapunov Bounds and E cient Importance Sampling for a Heavy-tailed G/G/1 Queue Fluid Heuristics, Lyapunov Bounds and E cient Importance Sampling for a Heavy-tailed G/G/1 Queue J. Blanchet, P. Glynn, and J. C. Liu. September, 2007 Abstract We develop a strongly e cient rare-event

More information

ABC methods for phase-type distributions with applications in insurance risk problems

ABC methods for phase-type distributions with applications in insurance risk problems ABC methods for phase-type with applications problems Concepcion Ausin, Department of Statistics, Universidad Carlos III de Madrid Joint work with: Pedro Galeano, Universidad Carlos III de Madrid Simon

More information

E cient Simulation and Conditional Functional Limit Theorems for Ruinous Heavy-tailed Random Walks

E cient Simulation and Conditional Functional Limit Theorems for Ruinous Heavy-tailed Random Walks E cient Simulation and Conditional Functional Limit Theorems for Ruinous Heavy-tailed Random Walks Jose Blanchet and Jingchen Liu y June 14, 21 Abstract The contribution of this paper is to introduce change

More information

Introduction to Algorithmic Trading Strategies Lecture 10

Introduction to Algorithmic Trading Strategies Lecture 10 Introduction to Algorithmic Trading Strategies Lecture 10 Risk Management Haksun Li haksun.li@numericalmethod.com www.numericalmethod.com Outline Value at Risk (VaR) Extreme Value Theory (EVT) References

More information

Eco517 Fall 2004 C. Sims MIDTERM EXAM

Eco517 Fall 2004 C. Sims MIDTERM EXAM Eco517 Fall 2004 C. Sims MIDTERM EXAM Answer all four questions. Each is worth 23 points. Do not devote disproportionate time to any one question unless you have answered all the others. (1) We are considering

More information

Poisson Processes. Stochastic Processes. Feb UC3M

Poisson Processes. Stochastic Processes. Feb UC3M Poisson Processes Stochastic Processes UC3M Feb. 2012 Exponential random variables A random variable T has exponential distribution with rate λ > 0 if its probability density function can been written

More information

Quantifying Stochastic Model Errors via Robust Optimization

Quantifying Stochastic Model Errors via Robust Optimization Quantifying Stochastic Model Errors via Robust Optimization IPAM Workshop on Uncertainty Quantification for Multiscale Stochastic Systems and Applications Jan 19, 2016 Henry Lam Industrial & Operations

More information

Severity Models - Special Families of Distributions

Severity Models - Special Families of Distributions Severity Models - Special Families of Distributions Sections 5.3-5.4 Stat 477 - Loss Models Sections 5.3-5.4 (Stat 477) Claim Severity Models Brian Hartman - BYU 1 / 1 Introduction Introduction Given that

More information

Introduction and Overview STAT 421, SP Course Instructor

Introduction and Overview STAT 421, SP Course Instructor Introduction and Overview STAT 421, SP 212 Prof. Prem K. Goel Mon, Wed, Fri 3:3PM 4:48PM Postle Hall 118 Course Instructor Prof. Goel, Prem E mail: goel.1@osu.edu Office: CH 24C (Cockins Hall) Phone: 614

More information

Experience Rating in General Insurance by Credibility Estimation

Experience Rating in General Insurance by Credibility Estimation Experience Rating in General Insurance by Credibility Estimation Xian Zhou Department of Applied Finance and Actuarial Studies Macquarie University, Sydney, Australia Abstract This work presents a new

More information

Theory and Applications of Stochastic Systems Lecture Exponential Martingale for Random Walk

Theory and Applications of Stochastic Systems Lecture Exponential Martingale for Random Walk Instructor: Victor F. Araman December 4, 2003 Theory and Applications of Stochastic Systems Lecture 0 B60.432.0 Exponential Martingale for Random Walk Let (S n : n 0) be a random walk with i.i.d. increments

More information

EFFICIENT SIMULATION FOR LARGE DEVIATION PROBABILITIES OF SUMS OF HEAVY-TAILED INCREMENTS

EFFICIENT SIMULATION FOR LARGE DEVIATION PROBABILITIES OF SUMS OF HEAVY-TAILED INCREMENTS Proceedings of the 2006 Winter Simulation Conference L. F. Perrone, F. P. Wieland, J. Liu, B. G. Lawson, D. M. Nicol, and R. M. Fujimoto, eds. EFFICIENT SIMULATION FOR LARGE DEVIATION PROBABILITIES OF

More information

EXACT SAMPLING OF THE INFINITE HORIZON MAXIMUM OF A RANDOM WALK OVER A NON-LINEAR BOUNDARY

EXACT SAMPLING OF THE INFINITE HORIZON MAXIMUM OF A RANDOM WALK OVER A NON-LINEAR BOUNDARY Applied Probability Trust EXACT SAMPLING OF THE INFINITE HORIZON MAXIMUM OF A RANDOM WALK OVER A NON-LINEAR BOUNDARY JOSE BLANCHET, Columbia University JING DONG, Northwestern University ZHIPENG LIU, Columbia

More information

IEOR 4106: Spring Solutions to Homework Assignment 7: Due on Tuesday, March 22.

IEOR 4106: Spring Solutions to Homework Assignment 7: Due on Tuesday, March 22. IEOR 46: Spring Solutions to Homework Assignment 7: Due on Tuesday, March. More of Chapter 5: Read the rest of Section 5.3, skipping Examples 5.7 (Coupon Collecting), 5. (Insurance claims)and Subsection

More information

A polynomial expansion to approximate ruin probabilities

A polynomial expansion to approximate ruin probabilities A polynomial expansion to approximate ruin probabilities P.O. Goffard 1 X. Guerrault 2 S. Loisel 3 D. Pommerêt 4 1 Axa France - Institut de mathématiques de Luminy Université de Aix-Marseille 2 Axa France

More information

Applying the proportional hazard premium calculation principle

Applying the proportional hazard premium calculation principle Applying the proportional hazard premium calculation principle Maria de Lourdes Centeno and João Andrade e Silva CEMAPRE, ISEG, Technical University of Lisbon, Rua do Quelhas, 2, 12 781 Lisbon, Portugal

More information

Poisson Approximation for Two Scan Statistics with Rates of Convergence

Poisson Approximation for Two Scan Statistics with Rates of Convergence Poisson Approximation for Two Scan Statistics with Rates of Convergence Xiao Fang (Joint work with David Siegmund) National University of Singapore May 28, 2015 Outline The first scan statistic The second

More information

2WB05 Simulation Lecture 7: Output analysis

2WB05 Simulation Lecture 7: Output analysis 2WB05 Simulation Lecture 7: Output analysis Marko Boon http://www.win.tue.nl/courses/2wb05 December 17, 2012 Outline 2/33 Output analysis of a simulation Confidence intervals Warm-up interval Common random

More information

Weak quenched limiting distributions of a one-dimensional random walk in a random environment

Weak quenched limiting distributions of a one-dimensional random walk in a random environment Weak quenched limiting distributions of a one-dimensional random walk in a random environment Jonathon Peterson Cornell University Department of Mathematics Joint work with Gennady Samorodnitsky September

More information

Distributed Estimation, Information Loss and Exponential Families. Qiang Liu Department of Computer Science Dartmouth College

Distributed Estimation, Information Loss and Exponential Families. Qiang Liu Department of Computer Science Dartmouth College Distributed Estimation, Information Loss and Exponential Families Qiang Liu Department of Computer Science Dartmouth College Statistical Learning / Estimation Learning generative models from data Topic

More information

Multivariate Risk Processes with Interacting Intensities

Multivariate Risk Processes with Interacting Intensities Multivariate Risk Processes with Interacting Intensities Nicole Bäuerle (joint work with Rudolf Grübel) Luminy, April 2010 Outline Multivariate pure birth processes Multivariate Risk Processes Fluid Limits

More information

Mathematical statistics

Mathematical statistics October 4 th, 2018 Lecture 12: Information Where are we? Week 1 Week 2 Week 4 Week 7 Week 10 Week 14 Probability reviews Chapter 6: Statistics and Sampling Distributions Chapter 7: Point Estimation Chapter

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

If we want to analyze experimental or simulated data we might encounter the following tasks:

If we want to analyze experimental or simulated data we might encounter the following tasks: Chapter 1 Introduction If we want to analyze experimental or simulated data we might encounter the following tasks: Characterization of the source of the signal and diagnosis Studying dependencies Prediction

More information

IEOR E4703: Monte-Carlo Simulation

IEOR E4703: Monte-Carlo Simulation IEOR E4703: Monte-Carlo Simulation Output Analysis for Monte-Carlo Martin Haugh Department of Industrial Engineering and Operations Research Columbia University Email: martin.b.haugh@gmail.com Output Analysis

More information

Terminology Suppose we have N observations {x(n)} N 1. Estimators as Random Variables. {x(n)} N 1

Terminology Suppose we have N observations {x(n)} N 1. Estimators as Random Variables. {x(n)} N 1 Estimation Theory Overview Properties Bias, Variance, and Mean Square Error Cramér-Rao lower bound Maximum likelihood Consistency Confidence intervals Properties of the mean estimator Properties of the

More information

Bayesian Semiparametric GARCH Models

Bayesian Semiparametric GARCH Models Bayesian Semiparametric GARCH Models Xibin (Bill) Zhang and Maxwell L. King Department of Econometrics and Business Statistics Faculty of Business and Economics xibin.zhang@monash.edu Quantitative Methods

More information

Bayesian Semiparametric GARCH Models

Bayesian Semiparametric GARCH Models Bayesian Semiparametric GARCH Models Xibin (Bill) Zhang and Maxwell L. King Department of Econometrics and Business Statistics Faculty of Business and Economics xibin.zhang@monash.edu Quantitative Methods

More information

A Central Limit Theorem for Fleming-Viot Particle Systems Application to the Adaptive Multilevel Splitting Algorithm

A Central Limit Theorem for Fleming-Viot Particle Systems Application to the Adaptive Multilevel Splitting Algorithm A Central Limit Theorem for Fleming-Viot Particle Systems Application to the Algorithm F. Cérou 1,2 B. Delyon 2 A. Guyader 3 M. Rousset 1,2 1 Inria Rennes Bretagne Atlantique 2 IRMAR, Université de Rennes

More information

Physics 403 Probability Distributions II: More Properties of PDFs and PMFs

Physics 403 Probability Distributions II: More Properties of PDFs and PMFs Physics 403 Probability Distributions II: More Properties of PDFs and PMFs Segev BenZvi Department of Physics and Astronomy University of Rochester Table of Contents 1 Last Time: Common Probability Distributions

More information

STA 2201/442 Assignment 2

STA 2201/442 Assignment 2 STA 2201/442 Assignment 2 1. This is about how to simulate from a continuous univariate distribution. Let the random variable X have a continuous distribution with density f X (x) and cumulative distribution

More information

A Note On The Erlang(λ, n) Risk Process

A Note On The Erlang(λ, n) Risk Process A Note On The Erlangλ, n) Risk Process Michael Bamberger and Mario V. Wüthrich Version from February 4, 2009 Abstract We consider the Erlangλ, n) risk process with i.i.d. exponentially distributed claims

More information

ACCOUNTING FOR INPUT-MODEL AND INPUT-PARAMETER UNCERTAINTIES IN SIMULATION. <www.ie.ncsu.edu/jwilson> May 22, 2006

ACCOUNTING FOR INPUT-MODEL AND INPUT-PARAMETER UNCERTAINTIES IN SIMULATION. <www.ie.ncsu.edu/jwilson> May 22, 2006 ACCOUNTING FOR INPUT-MODEL AND INPUT-PARAMETER UNCERTAINTIES IN SIMULATION Slide 1 Faker Zouaoui Sabre Holdings James R. Wilson NC State University May, 006 Slide From American

More information

March 10, 2017 THE EXPONENTIAL CLASS OF DISTRIBUTIONS

March 10, 2017 THE EXPONENTIAL CLASS OF DISTRIBUTIONS March 10, 2017 THE EXPONENTIAL CLASS OF DISTRIBUTIONS Abstract. We will introduce a class of distributions that will contain many of the discrete and continuous we are familiar with. This class will help

More information

f X (x) = λe λx, , x 0, k 0, λ > 0 Γ (k) f X (u)f X (z u)du

f X (x) = λe λx, , x 0, k 0, λ > 0 Γ (k) f X (u)f X (z u)du 11 COLLECTED PROBLEMS Do the following problems for coursework 1. Problems 11.4 and 11.5 constitute one exercise leading you through the basic ruin arguments. 2. Problems 11.1 through to 11.13 but excluding

More information

Pseudo-marginal MCMC methods for inference in latent variable models

Pseudo-marginal MCMC methods for inference in latent variable models Pseudo-marginal MCMC methods for inference in latent variable models Arnaud Doucet Department of Statistics, Oxford University Joint work with George Deligiannidis (Oxford) & Mike Pitt (Kings) MCQMC, 19/08/2016

More information

Paper Review: Risk Processes with Hawkes Claims Arrivals

Paper Review: Risk Processes with Hawkes Claims Arrivals Paper Review: Risk Processes with Hawkes Claims Arrivals Stabile, Torrisi (2010) Gabriela Zeller University Calgary, Alberta, Canada 'Hawks Seminar' Talk Dept. of Math. & Stat. Calgary, Canada May 30,

More information

IEOR 3106: Introduction to Operations Research: Stochastic Models. Professor Whitt. SOLUTIONS to Homework Assignment 2

IEOR 3106: Introduction to Operations Research: Stochastic Models. Professor Whitt. SOLUTIONS to Homework Assignment 2 IEOR 316: Introduction to Operations Research: Stochastic Models Professor Whitt SOLUTIONS to Homework Assignment 2 More Probability Review: In the Ross textbook, Introduction to Probability Models, read

More information

Uniformly Efficient Importance Sampling for the Tail Distribution of Sums of Random Variables

Uniformly Efficient Importance Sampling for the Tail Distribution of Sums of Random Variables Uniformly Efficient Importance Sampling for the Tail Distribution of Sums of Random Variables P. Glasserman Columbia University pg20@columbia.edu S.K. Juneja Tata Institute of Fundamental Research juneja@tifr.res.in

More information

Estimation for the standard and geometric telegraph process. Stefano M. Iacus. (SAPS VI, Le Mans 21-March-2007)

Estimation for the standard and geometric telegraph process. Stefano M. Iacus. (SAPS VI, Le Mans 21-March-2007) Estimation for the standard and geometric telegraph process Stefano M. Iacus University of Milan(Italy) (SAPS VI, Le Mans 21-March-2007) 1 1. Telegraph process Consider a particle moving on the real line

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

LECTURE 5 NOTES. n t. t Γ(a)Γ(b) pt+a 1 (1 p) n t+b 1. The marginal density of t is. Γ(t + a)γ(n t + b) Γ(n + a + b)

LECTURE 5 NOTES. n t. t Γ(a)Γ(b) pt+a 1 (1 p) n t+b 1. The marginal density of t is. Γ(t + a)γ(n t + b) Γ(n + a + b) LECTURE 5 NOTES 1. Bayesian point estimators. In the conventional (frequentist) approach to statistical inference, the parameter θ Θ is considered a fixed quantity. In the Bayesian approach, it is considered

More information

Bayesian Inference. Chapter 2: Conjugate models

Bayesian Inference. Chapter 2: Conjugate models Bayesian Inference Chapter 2: Conjugate models Conchi Ausín and Mike Wiper Department of Statistics Universidad Carlos III de Madrid Master in Business Administration and Quantitative Methods Master in

More information

Optimal Transport in Risk Analysis

Optimal Transport in Risk Analysis Optimal Transport in Risk Analysis Jose Blanchet (based on work with Y. Kang and K. Murthy) Stanford University (Management Science and Engineering), and Columbia University (Department of Statistics and

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

STA205 Probability: Week 8 R. Wolpert

STA205 Probability: Week 8 R. Wolpert INFINITE COIN-TOSS AND THE LAWS OF LARGE NUMBERS The traditional interpretation of the probability of an event E is its asymptotic frequency: the limit as n of the fraction of n repeated, similar, and

More information

Fall 2017 STAT 532 Homework Peter Hoff. 1. Let P be a probability measure on a collection of sets A.

Fall 2017 STAT 532 Homework Peter Hoff. 1. Let P be a probability measure on a collection of sets A. 1. Let P be a probability measure on a collection of sets A. (a) For each n N, let H n be a set in A such that H n H n+1. Show that P (H n ) monotonically converges to P ( k=1 H k) as n. (b) For each n

More information

Analysis of the ruin probability using Laplace transforms and Karamata Tauberian theorems

Analysis of the ruin probability using Laplace transforms and Karamata Tauberian theorems Analysis of the ruin probability using Laplace transforms and Karamata Tauberian theorems Corina Constantinescu and Enrique Thomann Abstract The classical result of Cramer-Lundberg states that if the rate

More information

Reinsurance and ruin problem: asymptotics in the case of heavy-tailed claims

Reinsurance and ruin problem: asymptotics in the case of heavy-tailed claims Reinsurance and ruin problem: asymptotics in the case of heavy-tailed claims Serguei Foss Heriot-Watt University, Edinburgh Karlovasi, Samos 3 June, 2010 (Joint work with Tomasz Rolski and Stan Zachary

More information

Extreme Value Analysis and Spatial Extremes

Extreme Value Analysis and Spatial Extremes Extreme Value Analysis and Department of Statistics Purdue University 11/07/2013 Outline Motivation 1 Motivation 2 Extreme Value Theorem and 3 Bayesian Hierarchical Models Copula Models Max-stable Models

More information

Zig-Zag Monte Carlo. Delft University of Technology. Joris Bierkens February 7, 2017

Zig-Zag Monte Carlo. Delft University of Technology. Joris Bierkens February 7, 2017 Zig-Zag Monte Carlo Delft University of Technology Joris Bierkens February 7, 2017 Joris Bierkens (TU Delft) Zig-Zag Monte Carlo February 7, 2017 1 / 33 Acknowledgements Collaborators Andrew Duncan Paul

More information

Stochastic Simulation

Stochastic Simulation Stochastic Simulation Jan-Pieter Dorsman 1 & Michel Mandjes 1,2,3 1 Korteweg-de Vries Institute for Mathematics, University of Amsterdam 2 CWI, Amsterdam 3 Eurandom, Eindhoven University of Amsterdam,

More information

Scientific Computing: Monte Carlo

Scientific Computing: Monte Carlo Scientific Computing: Monte Carlo Aleksandar Donev Courant Institute, NYU 1 donev@courant.nyu.edu 1 Course MATH-GA.2043 or CSCI-GA.2112, Spring 2012 April 5th and 12th, 2012 A. Donev (Courant Institute)

More information

1 of 7 7/16/2009 6:12 AM Virtual Laboratories > 7. Point Estimation > 1 2 3 4 5 6 1. Estimators The Basic Statistical Model As usual, our starting point is a random experiment with an underlying sample

More information

Rare Events in Random Walks and Queueing Networks in the Presence of Heavy-Tailed Distributions

Rare Events in Random Walks and Queueing Networks in the Presence of Heavy-Tailed Distributions Rare Events in Random Walks and Queueing Networks in the Presence of Heavy-Tailed Distributions Sergey Foss Heriot-Watt University, Edinburgh and Institute of Mathematics, Novosibirsk The University of

More information

Modèles de dépendance entre temps inter-sinistres et montants de sinistre en théorie de la ruine

Modèles de dépendance entre temps inter-sinistres et montants de sinistre en théorie de la ruine Séminaire de Statistiques de l'irma Modèles de dépendance entre temps inter-sinistres et montants de sinistre en théorie de la ruine Romain Biard LMB, Université de Franche-Comté en collaboration avec

More information

Market Risk. MFM Practitioner Module: Quantitiative Risk Management. John Dodson. February 8, Market Risk. John Dodson.

Market Risk. MFM Practitioner Module: Quantitiative Risk Management. John Dodson. February 8, Market Risk. John Dodson. MFM Practitioner Module: Quantitiative Risk Management February 8, 2017 This week s material ties together our discussion going back to the beginning of the fall term about risk measures based on the (single-period)

More information

Brandon C. Kelly (Harvard Smithsonian Center for Astrophysics)

Brandon C. Kelly (Harvard Smithsonian Center for Astrophysics) Brandon C. Kelly (Harvard Smithsonian Center for Astrophysics) Probability quantifies randomness and uncertainty How do I estimate the normalization and logarithmic slope of a X ray continuum, assuming

More information

Negative Association, Ordering and Convergence of Resampling Methods

Negative Association, Ordering and Convergence of Resampling Methods Negative Association, Ordering and Convergence of Resampling Methods Nicolas Chopin ENSAE, Paristech (Joint work with Mathieu Gerber and Nick Whiteley, University of Bristol) Resampling schemes: Informal

More information

RMSC 2001 Introduction to Risk Management

RMSC 2001 Introduction to Risk Management RMSC 2001 Introduction to Risk Management Tutorial 4 (2011/12) 1 February 20, 2012 Outline: 1. Failure Time 2. Loss Frequency 3. Loss Severity 4. Aggregate Claim ====================================================

More information

1 Review of Probability

1 Review of Probability 1 Review of Probability Random variables are denoted by X, Y, Z, etc. The cumulative distribution function (c.d.f.) of a random variable X is denoted by F (x) = P (X x), < x

More information

Statistics Ph.D. Qualifying Exam: Part I October 18, 2003

Statistics Ph.D. Qualifying Exam: Part I October 18, 2003 Statistics Ph.D. Qualifying Exam: Part I October 18, 2003 Student Name: 1. Answer 8 out of 12 problems. Mark the problems you selected in the following table. 1 2 3 4 5 6 7 8 9 10 11 12 2. Write your answer

More information

Qualifying Exam in Probability and Statistics. https://www.soa.org/files/edu/edu-exam-p-sample-quest.pdf

Qualifying Exam in Probability and Statistics. https://www.soa.org/files/edu/edu-exam-p-sample-quest.pdf Part : Sample Problems for the Elementary Section of Qualifying Exam in Probability and Statistics https://www.soa.org/files/edu/edu-exam-p-sample-quest.pdf Part 2: Sample Problems for the Advanced Section

More information

Hypothesis Testing Problem. TMS-062: Lecture 5 Hypotheses Testing. Alternative Hypotheses. Test Statistic

Hypothesis Testing Problem. TMS-062: Lecture 5 Hypotheses Testing. Alternative Hypotheses. Test Statistic Hypothesis Testing Problem TMS-062: Lecture 5 Hypotheses Testing Same basic situation as befe: Data: random i. i. d. sample X 1,..., X n from a population and we wish to draw inference about unknown population

More information

Stat 516, Homework 1

Stat 516, Homework 1 Stat 516, Homework 1 Due date: October 7 1. Consider an urn with n distinct balls numbered 1,..., n. We sample balls from the urn with replacement. Let N be the number of draws until we encounter a ball

More information

Economics 520. Lecture Note 19: Hypothesis Testing via the Neyman-Pearson Lemma CB 8.1,

Economics 520. Lecture Note 19: Hypothesis Testing via the Neyman-Pearson Lemma CB 8.1, Economics 520 Lecture Note 9: Hypothesis Testing via the Neyman-Pearson Lemma CB 8., 8.3.-8.3.3 Uniformly Most Powerful Tests and the Neyman-Pearson Lemma Let s return to the hypothesis testing problem

More information

Stat 535 C - Statistical Computing & Monte Carlo Methods. Arnaud Doucet.

Stat 535 C - Statistical Computing & Monte Carlo Methods. Arnaud Doucet. Stat 535 C - Statistical Computing & Monte Carlo Methods Arnaud Doucet Email: arnaud@cs.ubc.ca 1 CS students: don t forget to re-register in CS-535D. Even if you just audit this course, please do register.

More information

Monte-Carlo MMD-MA, Université Paris-Dauphine. Xiaolu Tan

Monte-Carlo MMD-MA, Université Paris-Dauphine. Xiaolu Tan Monte-Carlo MMD-MA, Université Paris-Dauphine Xiaolu Tan tan@ceremade.dauphine.fr Septembre 2015 Contents 1 Introduction 1 1.1 The principle.................................. 1 1.2 The error analysis

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

Orthonormal polynomial expansions and lognormal sum densities

Orthonormal polynomial expansions and lognormal sum densities 1/28 Orthonormal polynomial expansions and lognormal sum densities Pierre-Olivier Goffard Université Libre de Bruxelles pierre-olivier.goffard@ulb.ac.be February 22, 2016 2/28 Introduction Motivations

More information

Approximating the Integrated Tail Distribution

Approximating the Integrated Tail Distribution Approximating the Integrated Tail Distribution Ants Kaasik & Kalev Pärna University of Tartu VIII Tartu Conference on Multivariate Statistics 26th of June, 2007 Motivating example Let W be a steady-state

More information

A LOGARITHMIC EFFICIENT ESTIMATOR OF THE PROBABILITY OF RUIN WITH RECUPERATION FOR SPECTRALLY NEGATIVE LEVY RISK PROCESSES.

A LOGARITHMIC EFFICIENT ESTIMATOR OF THE PROBABILITY OF RUIN WITH RECUPERATION FOR SPECTRALLY NEGATIVE LEVY RISK PROCESSES. A LOGARITHMIC EFFICIENT ESTIMATOR OF THE PROBABILITY OF RUIN WITH RECUPERATION FOR SPECTRALLY NEGATIVE LEVY RISK PROCESSES Riccardo Gatto Submitted: April 204 Revised: July 204 Abstract This article provides

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

University Of Calgary Department of Mathematics and Statistics

University Of Calgary Department of Mathematics and Statistics University Of Calgary Department of Mathematics and Statistics Hawkes Seminar May 23, 2018 1 / 46 Some Problems in Insurance and Reinsurance Mohamed Badaoui Department of Electrical Engineering National

More information

Part IA Probability. Definitions. Based on lectures by R. Weber Notes taken by Dexter Chua. Lent 2015

Part IA Probability. Definitions. Based on lectures by R. Weber Notes taken by Dexter Chua. Lent 2015 Part IA Probability Definitions Based on lectures by R. Weber Notes taken by Dexter Chua Lent 2015 These notes are not endorsed by the lecturers, and I have modified them (often significantly) after lectures.

More information

Lecture 7 Introduction to Statistical Decision Theory

Lecture 7 Introduction to Statistical Decision Theory Lecture 7 Introduction to Statistical Decision Theory I-Hsiang Wang Department of Electrical Engineering National Taiwan University ihwang@ntu.edu.tw December 20, 2016 1 / 55 I-Hsiang Wang IT Lecture 7

More information

STAT 135 Lab 3 Asymptotic MLE and the Method of Moments

STAT 135 Lab 3 Asymptotic MLE and the Method of Moments STAT 135 Lab 3 Asymptotic MLE and the Method of Moments Rebecca Barter February 9, 2015 Maximum likelihood estimation (a reminder) Maximum likelihood estimation Suppose that we have a sample, X 1, X 2,...,

More information

System Identification, Lecture 4

System Identification, Lecture 4 System Identification, Lecture 4 Kristiaan Pelckmans (IT/UU, 2338) Course code: 1RT880, Report code: 61800 - Spring 2012 F, FRI Uppsala University, Information Technology 30 Januari 2012 SI-2012 K. Pelckmans

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

Lecture 3. G. Cowan. Lecture 3 page 1. Lectures on Statistical Data Analysis

Lecture 3. G. Cowan. Lecture 3 page 1. Lectures on Statistical Data Analysis Lecture 3 1 Probability (90 min.) Definition, Bayes theorem, probability densities and their properties, catalogue of pdfs, Monte Carlo 2 Statistical tests (90 min.) general concepts, test statistics,

More information

System Identification, Lecture 4

System Identification, Lecture 4 System Identification, Lecture 4 Kristiaan Pelckmans (IT/UU, 2338) Course code: 1RT880, Report code: 61800 - Spring 2016 F, FRI Uppsala University, Information Technology 13 April 2016 SI-2016 K. Pelckmans

More information

RMSC 2001 Introduction to Risk Management

RMSC 2001 Introduction to Risk Management RMSC 2001 Introduction to Risk Management Tutorial 4 (2011/12) 1 February 20, 2012 Outline: 1. Failure Time 2. Loss Frequency 3. Loss Severity 4. Aggregate Claim ====================================================

More information