This action might not be possible to undo. Are you sure you want to continue?

BooksAudiobooksComicsSheet Music### Categories

### Categories

### Categories

Editors' Picks Books

Hand-picked favorites from

our editors

our editors

Editors' Picks Audiobooks

Hand-picked favorites from

our editors

our editors

Editors' Picks Comics

Hand-picked favorites from

our editors

our editors

Editors' Picks Sheet Music

Hand-picked favorites from

our editors

our editors

Top Books

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Audiobooks

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Comics

What's trending, bestsellers,

award-winners & more

award-winners & more

Top Sheet Music

What's trending, bestsellers,

award-winners & more

award-winners & more

Welcome to Scribd! Start your free trial and access books, documents and more.Find out more

1. (a) i. λ1 = λ, λ2 = 2λ, . . . , λn = nλ, . . .. (When there are n individuals at time t, i.e., X (t) = n, the probability of an arrival in the time interval (t, t + h] is nλh + o(h).) Note that this is just the “Yule process” from the lectures. ii. T1 = 0, so T5 = T1 + S2 + S3 + S4 + S5 = S2 + S3 + S4 + S5 . Now the interarrival time Sn is distributed Exp(λn−1 ), so E(T5 | T1 = 0) = E(S2 + S3 + S4 + S5 ) = E(S2 ) + E(S3 ) + E(S4 ) + E(S5 ) 1 1 1 25 1 + + = . = + λ 2λ 3λ 4λ 12λ iii.

∞

n=1

1 = λn

∞

n=1

1 1 = nλ λ

∞

n=1

1 = ∞, n

so the distribution of X (t) is deﬁned for all t. (b) i. λ2 = λ, λ3 = 3 λ, λ4 = 4 λ, . . . , λn = n λ, . . .. (When there 2 2 2 are n individuals at time t, i.e., X (t) = n, there are n unordered 2 pairs, and the probability of an arrival in the time interval (t, t + h] is n λh + o(h).) 2 ii. T2 = 0, so T5 = T2 + S3 + S4 + S5 = S3 + S4 + S5 . Now the interarrival time Sn is distributed Exp(λn−1 ), so E(T5 | T2 = 0) = E(S3 + S4 + S5 ) = E(S3 ) + E(S4 ) + E(S5 ) 1 1 3 1 . = + 3 + 4 = λ 2λ λ λ 2 2 iii.

∞

n=2

1 = λn

∞

1

n 2

n=2

λ

=

2 λ

∞

n=2

1 2 = n(n − 1) λ

∞

n=2

1 1 − n−1 n

=

2 , λ

so the distribution of X (t) is not deﬁned (for any t). (The process “explodes” at ﬁnite time. The process is not biologically plausible, so we shouldn’t be surprised!) 1

(a) The equations are pn (t) = −nλpn (t) + (n − 1)λpn−1 (t) for n ≥ 1. It follows that • P(X (t + h) = n|X (t) = n) = P(no arrivals) P(no births) = (1 − αh + o(h)) (1 − βh + o(h))n = 1 − αh − nβh + o(h) • P(X (t+h) = n + 1|X (t) = n) = P(1 arrival) P(no births) + P(no arrivals) P(1 birth) = (αh + o(h)) (1 − βh + o(h))n + (1 − αh + o(h)) n (βh + o(h)) (1 − βh + o(h))n−1 = αh + nβh + o(h). 3]. (b) If X (t) = n there are n creatures each of which produces oﬀspring according to a Poisson process of rate β The arrivals generate creatures according to a Poisson process of rate α.2. Since the arrivals form a Poisson process of rate α the probability of this is e−3α . 2 . • For the population to increase by 2 or more we must have more than one arrival or more than one birth or one arrival and one birth. (c) We have that P(X (t) = n) = p(1 − p)n−1 where p = e−λt and so X (t) ∼ Geom(e−λt ). (b) If we set pn (t) = e−λt 1 − e−λt pn (t) = −λe−λt 1 − e−λt = −λe−λt 1 − e−λt n−1 n−1 n−1 then n−2 n−2 + e−λt λe−λt (n − 1) 1 − e−λt n−2 n−1 − 1 − e−λt e−λt λ(n − 1) 1 − e−λt n−2 + e−λt λ(n − 1) 1 − e−λt = e−λt λ(n − 1) 1 − e−λt − λne−λt 1 − e−λt = (n − 1)λpn−1 (t) − nλpn (t) and so this is indeed a solution to the equations. (a) The island will be uninhabited at time 3 if and only if there are no arrivals in the interval [0. Hence E(X (t)) = eλt and Var(X (t)) = 1 − e−λt e2λt 3. Hence the probability that the island is inhabited after 3 months is 1 − e−3α .

Now P(more that 1 arrival) = o(h) and P(more than 1 birth) = b≥2 n (βh)b (1 − βh)n−b + o(h) = o(h). (c) In lectures we derived the following diﬀerential equations: p0 (t) = −λ0 p0 (t) p1 (t) = λ0 p0 (t) − λ1 p1 (t) p2 (t) = λ1 p1 (t) − λ2 p2 (t). p0 (t) = e−t . λ1 = 3. The method of that proof can be used to solve them. 2 2 Similarly. And so (using the initial condition p1 (0) = 0). where λk are as above. b Finally P(1 arrival and 1 birth) = (αh + o(h)) (nβh + o(h)) = o(h). Rearranging the second equation gives p1 (t)eλ1 t = eλ1 t λ0 p0 (t) = e2t . We know that p0 (0) = 1 and so C = 1. 2 2 3 . λ2 = 5. It follows that (α + nβ ) h + o(h) 1 − (α + nβ ) h + o(h) P(X (t + h) = m|X (t) = n) = o(h) 0 if if if if m=n+1 m=n m>n+1 m<n Since arrivals and births for each individual after time t are independent of the process up to t (because each is a Poisson process) we do have that for s > t X (s) − X (t) is independent of the process up to t. it is clear that the ﬁrst equation has solution p0 (t) = Ce−t . Firstly. rearranging the third equation p2 (t)eλ2 t 3 3 = eλ2 t λ1 p1 (t) = e4t − e2t . (d) We have λ0 = 1. In lectures we showed that these diﬀerential equations always have a unique the solution. Hence X (t) is a birth process with parameters λn = α + nβ . t p1 (t) = e −λ1 t 0 1 1 e2x dx = e−t − e−3t . Hence.

e. 2 2 8 4 8 (e) The required probability is p2 (3) which is 3 −3 3 −9 3 −15 e − e + e = 0. We know that Si ∼ Exp(λi−1 ) and so E(Si ) = λ− i−1 . b(t) = (p00 (t) − p01 (t)) = p00 (t) − p01 (t) = −2p00 (t) + 2p01 (t) = −2b(t). We want E(S1 +S2 +· · ·+S10 ) = 4. t p2 (t) = e−λ2 t 0 3 4x 3 2x 3 3 3 e − e dx = e−t − e−3t + e−5t . Solving. 2 1 1 1 1 1 1 1 + +· · ·+ = + + +· · ·+ = 2. (a(t) + b(t)) = 1 (1 + e−2t ) and Combining these two solutions. (f) The naive way to approach this would be to calculate p10 (t) using the above method. A better way is to deﬁne Si to be the time between the population size reaching i − 1 and it the population 1 size reaching i. the boundary condition implies D = 1. i. so a(t) = C . However. p00 (t) = 1 2 2 1 p01 (t) = 1 (a(t) − b(t)) = 2 (1 − e−2t ).And so.019 · · · 8 4 8 (if you calculated the numerical value here you will have noticed that all but the ﬁrst of the terms are negligible). According to the boundary condition. and p00 (t) + p01 (t) = a(t) = 1. (c) Adding the equations from (a). this will be time consuming as we only know how to calculate the pn (t) one by one. b(t) = De−2t . Subtracting the equations from (a). C = 1. (a) p00 (t) = −λ0 p00 (t) + µ1 p01 (t) = −p00 (t) + p01 (t). p01 (t) = +λ0 p00 (t) − µ1 p01 (t) = +p00 (t) − p01 (t) (b) The boundary conditions are p00 (0) = 1 and p01 (0) = 0. p00 (t) − p01 (t) = b(t) = e−2t .. Again. a(t) = (p00 (t) + p01 (t)) = p00 (t) + p01 (t) = 0.13 · · · λ0 λ1 λ9 1 3 5 19 4 . (using the initial condition p2 (0) = 0).

s

s

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue listening from where you left off, or restart the preview.

scribd