Simple Random Walk. Timo Leenman, June 3, Bachelor scription Supervisor: Prof. dr. F den Hollander

Size: px
Start display at page:

Download "Simple Random Walk. Timo Leenman, June 3, Bachelor scription Supervisor: Prof. dr. F den Hollander"

Transcription

1 Simple Radom Walk Timo Leema, Jue 3, 2008 Bachelor scriptio Supervisor: Prof. dr. F de Hollader

2 Cotets Defiitio of the radom walk 3 2 Recurrece of the radom walk 3 3 Rage of the radom walk 0 4 Probability measures ad stochastic covergece 5 5 Browia motio 8 Preface This treatise is o simple radom walk, ad o the way it gives rise to Browia motio. It was writte as my bachelor project, ad it was writte i such a way that it should serve as a good itroductio ito the subject for studets that have as much kowledge as I whe I bega workig o it. That is: a basic probability course, ad a little bit of measure theory. To that ed, the followig track is followed: I sectio, the simple radom walk is defied. I sectio 2, the first major limit property is studied: whether the walk be recurret or ot. Some calculus ad the discrete Fourier trasform are required to prove the result. I sectio 3, a secod limit property is studied: its rage, or, the umber of visited sites. I the full proof of the results, the otio of strog ad weak covergece presets itself, ad also the otio of tail evets. To uderstad these problems more precisely, ad as a ecessary preparatio for Browia motio, some measure theoretic foudatios are treated i sectio 4. Emphasis is put, ot o the formal derivatio of the results, but o the right otio of them i our cotext. I sectio 5, Browia motio is studied. First, i what maer simple radom walk gives rise to it, ad secodly its formal defiitio. Special care is devoted to explai the exact steps that are eeded for its costructio, for that is somethig which I foud rather difficult to uderstad from the texts I read o it. Timo Leema 2

3 Defiitio of the radom walk Radom walk describes the motio o a lattice, say Z d, of a walker that jumps at discrete time steps t, 2, 3,... to a ew, radomly chose, site. Such a radom walk ca be defied i various ways, resultig i various properties. With each time a radom variable is associated (the step), ad the distributio of these radom variables fixes the behaviour of the walk. Oe could for example defie a walk that ever visits a site twice, or oe that ever turs 80 degrees at oce. But the radom walk we are to look at here is simple radom walk. Its successive steps are chose idepedetly, they ca be of legth oly ad are chose uiformly out of the 2d possible directios o Z d. Formally, defie steps X i as radom variables o Z d as follows: Defiitio. The discrete radom variables X, X 2,... o Z d are called steps of the radom walk ad have the followig probability distributio: i N : P (X i e) 2d if e Zd ad e, ad P (X i e) 0 otherwise. Defiitio. S 0 0 Z d ad S X X for N is called the positio of the radom walk at time. So the radom walker begis his walk o the startig site S 0 0, ad by takig i.i.d. steps X i arrives at positio S at time. Because of the idepedece of the steps it is obvious that {S } N0 is a Markov process o the state space Z d, sice the future positios oly deped o the curret positio. Now that we have a Markov chai, we ca talk about trasitio probabilities. Our otatio shall be thus: call p(l) P (X l) P (S l) the oe-step trasitio probability, which is 2d for l a eighbour, ad 0 otherwise. Call P (l) P (S l) the -step trasitio probability, which equals the probability that the walk is at site l at time (startig i 0). 2 Recurrece of the radom walk I studyig the log term behaviour of the radom walk, oe of the first questios oe might be iterested i is whether the radom walker returs to its startig site. To this ed, defie F to be the probability that the radom walker evetually returs to the startig site S 0. If F, the the site S 0 is called recurret, if F <, the it is called trasiet. I the recurret case, it is obvious that the radom walker returs ot oly oce but ifiitely may times to S 0, whereas i the trasiet case, the radom walker may ever retur with positive probability F. I which latter case the umber of returs to S 0 is geometrically distributed with parameter F, ad therefore the expected umber of returs to a trasiet startig site is F F F. 3

4 The state space of a geeral Markov chai ca be partitioed ito recurret ad trasiet classes of states. I the simple walk, however, it is clear that all states commuicate (i.e. the walker has positive probability to reach ay give site startig from ay other give site), ad hece that it cosists of oly oe class. Therefore it makes sese to call the whole radom walk recurret or trasiet, wheever S 0 is so. The certaity of returig to every visited site o the oe had, ad the likelihood of ot returig to them o the other had, give recurret ad trasiet radom walks completely differet behaviour. A valuable result, due to Polýa, tells which simple radom walks are recurret: Polýa s Theorem. Simple radom walks of dimesio d, 2 are recurret, ad of d 3 are trasiet. The proof of this theorem is the object of this sectio. To that ed, we first eed a geeral criterio to see whether S 0 be recurret or ot. Theorem. The state S 0 0 is trasiet if ad oly if P (0) <. Proof. For t N, defie I t if S t 0, ad I t 0 otherwise. Note that N t I t is the umber of times that S 0 is revisited. For the expectatio of that umber the followig holds: [ ] E[N] E I t E[I t ] P (S t 0) P t (0). t t Computig the expectatio agai, we have t t E[N] kp (N k) [kp (N k) kp (N k + )] k P (N k) k k k F k, where the last equatio follows from the fact that every retur occurs idepedetly with probability F. Puttig the results together we get P t (0) E[N] t F k, k which diverges if F, ad coverges if F <. I other words, the radom walk is recurret precisely whe t P t(0) is ifiity. It is this equivalece that we will use to prove Polýa s theorem. First, we compute P (0). After that we compute the sum over to see 4

5 whether it diverges. There is a way to proceed that covers all dimesios at oce, amely, by givig a itegral expressio for P (0), ad aalyzig this expressio for. We will eed this procedure for dimesios d 3, but to illustrate that the situatio for d, 2 is sigificatly easier, we will first carry out a simple computatio for d, 2. d, 2 d : Imagie the oe-dimesioal lattice Z lyig horizotally. Ay path that the radom walker follows ca be uiquely represeted by a ifiite sequece (llrlrrrl...) (l stadig for left, r for right). Coversely, ay such sequece represets a path. Next, ote that ay path from 0 to itself has a eve legth, cotaiig as may steps to the left as to the right. Therefore P 2+ (0) 0; P 2 (0) ( 2 ) ( ) ( ) 2 2 (2)!!(2 )! 2 2. Now substitute Stirlig s approximatio of the factorial! e 2π as to get P 2 (0) 22 2 e 2 4π 2 e 2 2π So the ifiite sum becomes P (0) P 2 (0) π [ + o()] 2 2 as. π π [ + o()] > π, ad we see that the oe-dimesioal radom walk is recurret. d 2: For oe two-dimesioal walk, defie two oe-dimesioal walks i the followig way: 5

6 Let S be the two-dimesioal positio. Defie for every the positios of two oe-dimesioal walks S ad S 2 by the orthogoal projectio of S o the respective axes ad 2. The steps of a radom walk are the differeces betwee two successive positios, ad the two-dimesioal step X i ca take the values orth, east, south, west. The followig table gives the relatio betwee X i, X i ad X 2 i : N E S W X X 2 From this table it is obvious that the distributio of X give X 2 is the same as the margial distributio of X, ad P (X ) P (X ) P (X 2 ) P (X 2 ) 2. So i this way ay two oe-dimesioal radom walks correspod precisely to oe two-dimesioal radom walk, ad the other way roud. Therefore, i d 2 we ca write: P 2 (0) P (S 2 0) P (S2 0)P (S2 2 0) ( ) 2 π π ad, because still P 2+ (0) 0, the sum over becomes P (0) P 2 (0) π [ + o()]. So the two-dimesioal radom walk is recurret as well. d 3 The geeral method to prove recurrece or trasiece eeds some more computatio. The whole method is based o the well-kow theorem for Markov chais due to Chapma ad Kolmogorov, which i our otatio takes the form: Theorem. I the above otatio, the followig holds for all l Z d : P + (l) l Z d p(l l )P (l ). () I words this states that the probability of travellig to l i + steps ca be foud by summig over the positios the walker ca occupy at time. It is clear that the statemet uses the traslatio ivariace of the walk. The theorem ca be see as describig the evolutio of the walk: a recurrece relatio that expresses higher-step trasitio probabilities i terms of lower-step trasitio probabilities. The oe-step trasitio probabilities are prescribed by the defiitio of the radom walk. As may ordiary differetial equatios ca be solved by applyig the Fourier trasform F to them, 6

7 i a like maer we will solve our recurrece relatio usig the discrete Fourier trasform F, the properties of which are very much the same as of the cotiuous oe. It takes the form F(P (l)) P (k) l Z d e il k P (l), k [ π, π) d. For ease we defie the structure fuctio λ(k) as the Fourier traform of the oe-step trasitio probabilities: λ(k) p(k) l eil k p(l). We ca ow trasform equatio (), ad we get P + (k) l e il k l p(l l )P (l ). Note that e il k e i(l l ) k e il k is a costat (ot depedig o l ) that ca be placed behid the summatio sig. Therefore P + (k) p(l l )e i(l l ) k e il k P (l ). l l Now call m l l, ad the above equals p(m)e im k e il k P (l ) λ(k) P (k) P + (k). l m+l The recurrece relatio has ow become easy to solve: we oly eed the iitial coditio. This is P 0 (l) δ 0l (at t 0 the walker must be i the origi), ad i Fourier trasform this coditio is P 0 (k) l eil k δ 0l (oly for l 0 does the delta-fuctio ot vaish). Substitutig this iitial coditio, we see that the solutio of () becomes P (k) λ(k). The iverse Fourier trasform has the form of the d-dimesioal itegral P (l) F ( P (k)) (2π) d... e il k P (k)dk. k [ π,π) d So the formal solutio (for ay dimesio) of the trasitio probabilities P is P (l) (2π) d... e il k λ(k) dk. k [ π,π) d Util ow the calculatio was purely formal. If we wat to test the walk for trasiece, we must sum the trasitio probabilities P (0)over. For that, we eed to kow λ(k) ad the evaluate the multiple itegral. Because we are oly iterested i whether the sum P (0) coverge or ot, it suffices for our purpose to approximate λ(k), ad the to determie the limitig behaviour of P (0) for : if it approaches 0 fast eough, the the sum will coverge, ad hece the radom walk will be trasiet. 7

8 So, what is λ(k) for our simple radom walk? The aswer follows from a short computatio. Write the vector k (k, k 2,..., k d ) T, ad fill i the fomula for the structure fuctio: λ(k) p(k) l e il k p(l) l: l e il k 2d. Here we use that the oe-step trasitio probability p(l) equals 2d if l is a uit vector, ad equals 0 otherwise. Deote these 2d uit vectors by e j ad e j, for j,..., d. The λ(k) 2d 2d d d j d j [ e ie j k + e i( e j k) ] 2d d j [cos(k j ) + cos( k j )] + i 2d d cos(k j ). j [ ] e ik j + e ik j d [si(k j ) + si( k j )] For the sake of approximatig this with a fuctio that is more easily itegrated over, recall the followig two Taylor expasios for x 0: j cos(x) x2 2! + h.o. ad ex + x x2 2! + h.o., where h.o. meas higher order terms. Substitute this ito the formula for λ(k) to obtai λ(k) d d j ( k2 j 2! + h.o.) 2d k 2 + h.o. e 2d k 2 for k 0. The itegral we are to calculate has ow become P (0) (2π) d... e il k e 2d k 2 dk for k 0 k [ π,π) d but this approximatio oly holds for small k. Yet this is all we eed, because we are iterested i the limitig behaviour as, i which case the secod factor i the itegrad clearly becomes very small, uless k is take to be so small as to compesate for the large. Thus it ca be see that the domiatig ( cotributio to the itegral is cotaied i the regio where k O ). I the limit this meas that k approaches 8

9 0, ad the value of the itegral is ot affected whe we itegrate over the whole of R d i stead of oly [ π, π) d. Hece P (0) (2π) d... e 2d k 2 dk for. R d Now observe that the itegrad oly depeds o the legth of k. This suggests a trasformatio to spherical coordiates, because the the itegrad oly depeds o the radius r k. The above itegral equals (2π) d 0 d dr Bd r e 2d r2 dr, where Br d is the volume of the ball of radius r i d dimesios. For computig Br d, defie the stepfuctio θ [0, ). The, for all d, Br d θ(r 2 x 2 )dx. R d Now substitute y x r, ad hece x 2 r 2 y 2, ad dx r d dy. This yields Br d θ(r 2 r 2 y 2 )r d dy r d θ( y 2 )dy ω d r d, R d R d where ω d represets the volume of the uit sphere i d dimesios. Because d dr Bd r dr d ω d, our asymptotic itegral expressio for P (0) becomes (2π) d dω d r d e 2d r2 dr. 0 To evaluate further, substitute x 2d r2. The dx dr d r, so rdr d dx, ad ( ) d 2 ( ) 2d 2d r d 2 x 2 d x 2 d, ad the boudaries 0 ad remai the same. Now the itegral has become, for, ) 2 d P (0) x 2 d e x d dx (2π) d 0 dω d ( 2d (2π) d ω d(2d) 2 d Γ ( ) 2 d C(d) 2 d, 2 d where Γ(a) 0 e x x a dx for a > 0, ad C(d) is a costat that depeds o the dimesio. At last we ca sum over the trasitio probabilities. Note that C(d) C(d) < if d 3. d 2 d 2 Because for large eough P (0) 2 C(d), we coclude that for d 3 2 d it must hold that P (0) <, ad therefore simple radom walk i dimesio d 3 is trasiet. 9

10 3 Rage of the radom walk Ituitively, it is clear that a trasiet radom walker is much more likely to visit ew sites of the lattice tha a recurret oe. To make this precise, defie the rage R of a radom walk at time as the umber of distict poits visited withi time : Defiitio. 0 : R card({0 S 0, S,..., S }). Defiig F to be the probability that the walker will evetually retur to its startig site S 0, the behaviour of the rage is stated i the followig theorem: Theorem. ɛ > 0 : lim P ( R ( F ) > ɛ) 0. Proof. Defie φ 0 ad, for k N φ k if S i S k for all i,.., k ad φ k 0 otherwise. I other words, φ k if ad oly if the radom walker visits a ew site o its k th step. It is obvious that R k0 φ k, ad because of liearity of expectatios also E[R ] k0 E[φ k] holds. For ay k N we ca write: E[φ k ] P (φ k ) P (S k S k, S k S k 2,..., S k S 0 0) P (S k S k 0, S k S k 2 0,..., S k 0) P (X k 0, X k + X k 0,..., X k X 0) P (X 0, X + X 2 0,..., X X k 0) P (S j 0 for j,..., k) k F j (0, 0), j where i the fourth lie we reverse the idices, ad i the sixth lie F j (0, 0) is the probability that the radom walker, startig i 0, returs to 0 for the first time o its j th step. Takig the limit k, we get lim E[φ k] F k ad this equals 0 if ad oly if the radom walk is recurret. Cosequetly, lim E[φ k ] lim E[R ] F. k0 To proceed, we cosider the recurret ad the trasiet case separately. For ɛ 0, write ( ) R P > ɛ P (R k) k ɛ P (R k) k:k>ɛ ɛ k:k>ɛ kp (R k) ɛ E[R ]. k0 0

11 For the recurret case it therefore follows that lim P ( R > ɛ) 0 for all ɛ > 0, ad we are doe. The trasiet case is more complicated. First, otice that for ay cotious radom variable X o R 0 the followig holds: E[X] 0 xp (X x)dx ɛ 0 xp (X x)dx + ɛ xp (X x)dx ɛ ɛ f(x)dx ɛp (X > ɛ) where f(x) is the prbability desity fuctio, ad this iequality also holds for discrete radom variables o R 0. Secodly, use this iequality for the radom variable R ( F ) 2. Startig from the probability we are iterested i, we get ( ) R P ( F ) > ɛ P ( R ( F ) 2 > 2 ɛ 2) 2 ɛ 2 E[ R ( F ) 2 ] 2 ɛ 2 E[R2 2( F )R + 2 ( F ) 2 ] 2 ɛ 2 E[R2 2R E[R ] + 2(E[R ]) 2 2( F )E[R ] + 2 ( F ) 2 ] 2 ɛ 2 E[R2 2R E[R ] + (E[R ]) 2 ] + 2 ɛ 2 (2 ( F ) 2 2( F )E[R ] + (E[R ]) 2 ) 2 ɛ 2 E[(R E[R ]) 2 ] + ( [ ]) 2 R ɛ 2 F E. Write E[(R E[R ]) 2 ] var(r ). To prove that the above probability coverges to zero as, it suffices to prove that lim var(r 2 ) 0, because it was already show that lim E[ R ] F, so that the secod term of the expressio vaishes i the limit. Cotiue by computig var(r ) as follows (usig the liearity of expectatios): var(r ) E[(R ) 2 (E[R ]) 2 ] E E 2 j0 k0 j0 k0 φ j φ k E j0 j0 φ j (E[φ j φ k ] E[φ j ]E[φ k ]) 0 j k φ j (E[φ j φ k ] E[φ j ]E[φ k ]) + k0 φ k [ ] E φ k k0 E j0 φ j E[φ j E[φ j ]]. j0 2

12 This last equality follows from the fact that summig over the elemets of a symmetric (square) matrix, oe may as well take twice the sum over the elemets uder the diagoal, ad add the diagoal elemets (otice that φ 2 j φ j). Because E[φ j E[φ j ]] E[φ j ], var(r ) ca be estimated by var(r ) 2 0 j k (E[φ j φ k ] E[φ j ]E[φ k ]) + E[φ j ]. But we ca estimate it by a yet simpler expressio: Notice that for 0 j < k, E[φ j φ k ] P (φ j φ k ) P (S j S α for 0 α < j, S k S β for 0 β < k) j0 P (S j S α for 0 α < j, S k S β for j < β < k) P (X j 0, X j + X j 0,..., X j X 0; X k 0, X k + X k 0,..., X k X j+ 0) P (X 0,..., X X j 0) P (X 0,..., X X k j 0). The factorizatio ad mixig of idices is allowed because the X i are i.i.d. Now recall that E[φ k ] P (X 0,..., X X k 0), so the iequality says that E[φ j φ k ] E[φ j ]E[φ k j ] for 0 j < k. Substitutio ito the former estimate of var(r ) yields var(r ) 2 E[φ j ] (E[φ k j E[φ k ]) + E[R ] j0 kj+ Sice E k [φ k ] k j F j(0, 0), we have {E[φ k ]} k0 is a mootoe oicreasig sequece. But for ay such sequece a a 2... a the sum (a k j a k ) (a + a a j ) (a j+ + a j a ) kj+ is maximized by takig j 2 (that is, roud off 2 dowward). Ideed, by takig j smaller, the left term icreases less tha the right oe ad, by takig j larger, the left term decreases more tha the right oe. Its maximum value is (a a 2 ) [(a a ) a a 2 ]. Takig a k E[φ k ], ad recallig that k0 E[φ k] E[R ], it therefore holds that var(r ) 2 j0 E[φ j ]E[R 2 + R 2 + R ] + E[R ]. Because we already showed that lim E[ j0 φ j] lim E[ R ] F, we get [ ] E[R lim 2 var(r ) 2( F ) lim E 2 + R 2 + R + 0 2

13 ( F 2( F ) 2 + F 2 ) ( F ) 0, which was still left to be proved. The above theorem states that the radom variable R coverges to F R i probability, but i fact a stroger statemet also holds: coverges to F almost surely. The proof thereof requires the ergodic theorem, which we caot prove here. The differece betwee these two types of covergece of radom variables is defied i sectio 4, but ituitively it meas the followig. Cosider the collectio of all possible paths (of ifiite legth) a radom walker might take. For each of those paths with every time a value R is associated. Strog covergece meas that, if oe of those paths is selected, the it is with probability oe a path for which the sequece ( R ) N coverges to F. If some arbitrary deviatio ɛ 0 is give, the the probability (whe selectig a path) p P ( R ( F ) > ɛ) depeds o. Covergece i probability meas that that the sequece (p ) N coverges to 0. Theorem. P (lim R F ). Proof. We will defie two sequeces (D ) N ad (R,M ) N such that D R R,M, ad use these to determie the value of the limit. First, defie R,M like R, but at every M th step forget which sites were already visited (but do ot reset the couter to 0). Formally, defie the rages of subsequet M-step walks ad ow add these up to get Z k (M) card{s km, S km+,..., S k(m+) }, R,M M k0 Z k (M), which is the sequece described above. It is obvious that R,M R. Note R that it is ot clear yet that lim exists at all, so that we will estimate dowwards by takig the limit iferior ad upwards by takig the limit superior (because these exist for every sequece). Thus it must hold that lim sup R lim sup M k0 Z k(m). 3

14 Now, the Z k (M) are i.i.d., ad so the strog law of large umbers ca be applied to obtai lim sup M M M k0 Z k (M) M E[Z 0(M)] M E[R M] a.s. Now take the limit M. We already saw that M E[R M] coverges to F, ad therefore we get lim sup R F a.s. Secodly, defie D as the umber of distict sites visited i time that the walker ever visits agai. Formally, let D k0 ψ k with ψ k if X k X k+i 0 for all i N ad ψ k 0 otherwise. So ψ k oly if after the k th step the walker ever returs to the site where it is at time k, i.e. if the walker visits the curret site for the last time. The umber of times that the walker visits a site for the last time is evidetly smaller tha the umber of sites visited, so D R. Y 0, Y,... is said to be a statioary sequece of radom variables if, for every k, the sequece Y k, Y k+,... has the same distributio, that is, for every, the ( + )-tuples (Y 0, Y,..., Y ) ad (Y k, Y k+,..., Y k+ ) have the same joit probability distributio. I particular, the Y i s are idetically distributed, but they may be depedet. Because of symmetry ad the Markov property of the simple radom walk, it is clear that (ψ k ) k N is a statioary sequece. Therefore, by the ergodic theorem the limit lim k0 D ψ k lim exists with probability (but it may still be a radom variable). But D lim assumig a certai value is a so-called tail evet. Ituitively, tail evets are those evets for a sequece of radom variables that would still have occurred if some fiite umber of those radom variables would have had a differet realisatio. Tail evets are treated more thoroughly i D sectio 4. Ideed, all evets of the form lim D < C, lim C etc., are tail evets, ad must therefore, accordig to Kolmogorov s 0- law, 4

15 occur with probability 0 or. Cosequetly, the limit must eeds be equal to a costat. This costat ca oly be the atural cadidate, amely, lim ψ k E[ψ 0 ] P (X X i 0 for all i ) F a.s. k0 Cosequetly, lim if R lim if D lim D F a.s. R Fially, ote that the first statemet (lim sup F a.s.) ad R the secod statemet (lim if F a.s.) together imply the R statemet of the theorem (lim F a.s.). 4 Probability measures ad stochastic covergece The purpose of this sectio is to defie more precisely what is meat by radom variables ad their covergece. This is doe i measure-theoretic terms, because that is the oly way to make precise our costructio of socalled Browia motio i sectio 5. While this treatise is o simple radom walk, ad ot o measure-theoretic probability, we will put more emphasis o the ituitive iterpretatio of the defiitios tha o the proof of their properties, which ca be foud i []. We use a probability space to model experimets ivolvig radomess. A probability space (Ω, Σ, P) is defied as follows: Sample space. Ω is a set, called the sample space, whereof the poits ω Ω are called sample poits. Evets. Σ is a σ-algebra of subsets of Ω, that is, a collectio of subsets with the property that, firstly, Ω Σ; secodly, wheever F Σ the F C Ω \ F Σ; thirdly, wheever F Σ( N), the F Σ. Notice that these imply that Σ cotais the empty set, ad is closed uder coutable itersectios. All F Σ are called measurable subsets, or (whe talkig about probability spaces) evets. Probability. P is called a probability measure o (Ω, Σ), that is, a fuctio P : Σ [0, ] that assigs to every evet a umber betwee 0 ad. P must satisfy: firstly, P( ) 0 ad P(Ω) ; secodly, wheever (F ) N is a sequece of disjoit evets with uio F F, the P(F ) N F (σ-additivity). 5

16 Radomess is cotaied i this model i the followig way: Whe performig a experimet, some ω Ω is chose i such a way that for every F Σ, P(F ) represets the probability that the chose sample poit ω belogs to F (i which case the evet F is said to occur). Some statemet S about the outcomes is said to be true almost surely (a.s.), or with probability, if F {ω : S(ω) is true} Σ ad P(F ). If R is a collectio of subsets of S, the σ(r), the σ-algebra geerated by R, is defied to be the smallest σ-algebra cotaied i Σ. The Borel σ-algebra B is the smallest σ-algebra that cotais all ope sets i R. A fuctio h : Ω R is called Σ-measurable if h : B Σ, that is, if h(a) Σ for all A B. (Compare cotiuous maps i topology: a map is called cotiuous if the iverse image F (G) is ope for all ope G. A map is called measurable if the iverse image of every measurable subset is measurable.) Give (Ω, Σ), a radom variable is a Σ-measurable fuctio. So, for a radom variable X: X : Ω R ad X : B Σ Give a collectio (Y γ ) γ C of maps Y γ : Ω R, its geerated σ-algebra Y σ(y γ : γ C) is defied to be the smallest σ-algebra Y o Ω such that each map Y γ is Y- measurable (that is, such that each map is a radom variable). The followig holds: Y σ(y γ : γ C) σ({ω Ω : Y γ (ω) B} : γ C, B B). If X is a radom variable for some (Ω, Σ), the obviously σ(x) Σ. Suppose (Ω, Σ, P) is a model for some experimet, ad that the experimet has bee performed, that is, some ω Ω has bee selected. Suppose further that (Y γ ) γ C is a collectio of radom variables associated with the experimet. Now cosider the values Y γ (ω), that is, the observed values (realisatios) of the radom variables. The the ituitive sigificace of the σ-algebra σ(y γ : γ C) is that it cosists precisely of those evets F for which it is possible to decide whether or ot F has occurred (i.e. whether or ot ω F ) o the basis of the values Y γ (ω) oly. Moreover, this must be possible for every ω Ω. Give a sequece of radom variables (X ) N ad the geerated σ- algebras T σ(x +, X +2,...), defie the tail σ-algebra T of the sequece (X ) N as follows: T N T. 6

17 So T cosists of those evets which ca be said to occur (or ot to occur) o the basis of the realisatios of the radom variables, beyod ay fiite idex. For such evets the followig theorem states, that they will either occur with certaity, or ot at all. Kolmogorov s 0- law. Let (X ) N be a sequece of idepedet radom variables, ad T the tail σ-algebra thereof. The F T P(F ) 0 or P(F ). Suppose X is a radom variable carried by the probability space (Ω, Σ, P). We have Ω X R [0, ] P Σ X B [0, ] P σ(x) X B. Defie the law L X of X by L X P X, so L X : B [0, ]. The law ca be show to be a probability measure o (R, B). The distributio fuctio of a radom variable X is a fuctio F X : R [0, ] defied by: F X (c) L x (, c) P(X c) P({ω : X(ω) c}). Because we have defied a radom variable as a fuctio from Ω to R, we ow have several otios of a covergig sequece of radom variables. The usual modes of covergece for fuctios are all well-defied for ay sequece (X ) N of radom variables, ad we may for example cosider uiform covergece or poitwise covergece to some radom variable X. The latter oe is weaker, ad we mea by it: ω Ω : lim (X (ω)) X(ω) or, shortly, X (ω) X(ω). But i practice, for radom variables we are oly iterested i yet weaker modes of covergece, which we defie below. A sequece (X ) N of radom variables is said to coverge to X: almost surely (or, with probability ), if P({ω Ω : X (ω) X(ω)}). Note that ot for all ω Ω eeds (X (ω)) N coverge to X(ω), but the set of ω s for which it does coverge, has probability oe. Which is the same as sayig that if for some radom ω Ω, the sequece of real umbers (X (ω)) N is cosidered, it is certai to coverge to X(ω). i probability, if ɛ > 0 : P({ω Ω : X (ω) X(ω) > ɛ}) 0. Note that (X (ω)) N may ot coverge to X(ω) for all ω Ω, but for ay ɛ > 0 the probability that X deviates from X more tha ɛ for a fixed teds to 0 as. 7

18 i distributio, if F X (c) F X (c) for all cotiuity poits c of F X. Which expresses othig but the poitwise covergece of the distributio fuctios, ad tells othig about the radom variables themselves. Covergece almost surely is also called strog covergece ad is deoted as, covergece i probability is also called weak covergece ad is deoted P, covergece i distributio is also called covergece i law ad is deoted. Note that for covergece i law, the sequece of radom variables eeds ot be defied o the same probability space as its limit. I particular, the X s may be defied o a discrete space, while X may be defied o a cotiuous space. For example, if P(X i ) for i, 2,..., the X X, with X uiformly distributed o [0, ]. It ca be show that strog covergece implies weak covergece, ad weak covergece implies covergece i law, but oe of the three are equivalet i geeral. 5 Browia motio Imagie the followig: istead of executig a radom walk o the stadard lattice with site distace at time steps of size, we make the lattice ever arrower ad reduce our time steps appropriately. Evetually we will get some radom process i cotiuous space ad time. Loosely speakig, makig the lattice arrower ad reducig the time steps should happe i some harmoized maer: whe the distace travelled per step teds to 0, the umber of steps per time uit should ted to ifiity i the right way to have the proportio of the visited area be costat. We proceed to show how the above idea ca be made precise. ( For t 0, cosider the sequece S t. This is a sequece ) N of positios, rescaled such that it coverges to a radom variable as, by the cetral limit theorem. To apply this theorem, we must kow expectatio ad variace of the steps X i. Obviously E[X i ] 0, ad so var(x i ) E[Xi 2] E[X i] 2 E[Xi 2] 2d 2d i 2. With E[X i ] µ ad var(x i ) σ 2 the cetral limit theorem states ad therefore, for µ 0 ad σ, i i X i µ σ N(0, ), X i i X i S N(0, ). 8

19 Now fix t 0 ad observe that S t S t t t S t N(0, ). t Multiplyig by t we ca ow see that S t must for t 0 coverge i distributio to the ormal distributio N(0, t) of expectatio 0 ad variace t. Because of the idepedece of the steps X i, it is ( also clear that for) ay t 0 ad s 0 with t s 0, it must hold that S t S s N(0, t s). Recall that covergece i distributio does ot mea covergece of the sequece of actual values whe a experimet is performed (i fact, the limit lim S t does ot exist with probability ). Therefore it is impossible to treat the desired cotiuous process as a actual limit of some rescaled radom walk, but the covergece i distributio strogly suggests a defiitio of the limitig process by usig the acquired ormal distributios, that is, a process with idepedet icremets that are ormally distributed. But first, let us defie precisely what we mea by a cotiuous process, ad state some of its properties. A stochastic process X is a parametrized collectio of radom variables (X t ) t T defied o a probability space (Ω, Σ, P ) ad assumig values i R d. For our process, T [0, ), ad hece it is called cotiuous. The fiite-dimesioal distributios of a cotiuous process X are the measures µ t,...,t k defied o (R d ) k by µ t,...,t k (B,..., B k ) P (X t B,..., X tk B k ), where B i are Borel sets, ad t i T, for i,..., k. I other words, the fiitedimesioal distributios are the joit laws of the fiite collectios of radom variables out of the cotiuous process. Properties like cotiuity of the paths of the process are therefore ot determied by the fiite-dimesioal distributios, ad hece it is clear that a process X is ot equivalet to its fiite dimesioal distributios. Coversely, give a set {µ t,...,t k : k N, t i T fori,..., k} of probability measures o (R d ) k, uder what coditios ca a stochastic process be costructed, that has µ t,...,t k as its fiite-dimesioal distributios? Sufficiet coditios are give i the followig theorem. Kolmogorov s extesio theorem. For t,..., t k T, let µ t,...,t k be probability measures o (R d ) k such that 9

20 ) µ tσ(),...,t σ(k) (B... B k ) µ t,...,t k (B σ ()... B σ (k)) for all permutatios σ o {, 2,..., k} ad all t i T, i,..., k; 2) µ t,...,t k (B,..., B k ) µ t,...,t k,t k+,...,t k+m (B... B k (R d ) m ) for all t i T, i,..., k; The there exists a probability space (Ω, Σ, P ), ad a cotiuous process (X t ) t T o Ω, such that for all E i R d, i,..., k: µ t,...,t k (E... E k ) P (X t E,..., X tk E k ). This gives us the existece of some process, whereof oly the fiite dimesioal distributios are kow. It tells us othig about the shape of Ω (which, of course, is ot uique), yet the theorem is of crucial importace, sice it allows us to cosider joit laws of ifiite collectios of radom variables draw out of the process, ad therefore questios o cotiuity of paths, etc. Our cosideratio of the rescalig of the radom walk yielded very atural ( cadidates for a collectio of measures o (R d ) k, amely, those to which S t S s coverge i probability. The procedure for the ) N formal defiitio of the desired process comprises therefore, firstly, the defiitio of such a collectio of probability measures o (R d ) k ad secodly, the applicatio of the extesio theorem. This is carried out below. Defie p(t, y) : R 0 R d R as the joit probability desity fuctio of d idepedet ormal radom variables with variace t 0: p(t, y) (2πt) d 2 e y 2 2t for y R d, t 0. I order to defie the required probability measures o (R d ) k, first defie for k N ad 0 t... t k the measure P t,...,t k by P t,...,t k (E... E k ) p(t, x )p(t 2 t, x 2 x )... p(t k t k, x k x k )dx...dx k, E... E k where as a covetio p(0, y)dy δ 0 to avoid icosistecies. Secodly, exted this defiitio to all fiite sequeces t,..., t k by usig the first coditio i Kolmogorov s extesio theorem. The also the secod coditio is satisfied, because p is a probability desity (that itegrates to ). So there exists a probability space (Ω, Σ, P ) ad a cotiuous process (B t ) t 0 o Ω such that the fiite-dimesioal distributios of B t are give by the prescribed oes. This process (B t ) t 0 is called Browia motio. The fact that it has idepedet ad ormally distributed icremets ca be easily see from our defiitio. The third essetial property of Browia motio is its cotiuity. To show this, we use aother theorem of Kolmogorov. 20

21 Kolmogorov s cotiuity theorem. Let X (X t ) t 0 be a cotiuoustime process. If for all T > 0 there exist α, β, D > 0 such that E [ X t X s α ] D t s +β for 0 s, t T, the the paths of X are cotiuous with probability, that is, P (t X t is cotiuous). We will use this result to show cotiuity of Browia motio i dimesio d. Because (B t B s ) N(0, t s), partial itegratio gives E [ B t B s α ] Take α 4 to get E[ B t B s 4 ] x α e t s 2π x2 2(t s) dx α t s 2π x α 2 e 2 2(t s) t s 2π α 2 2(t s) α 3 2 2(t s) t s 2π 3(t s)(t s) x2 2(t s) dx x α 4 e π ( 2(t s) 3(t s)2 t s 2π 2(t s)π 3(t s) 2. x2 2(t s) dx. Hece for α 4, D 3, β, Browia motio satisfies the cotiuity coditio E [ B t B s α ] D t s +β. Although the Browia motio as we defied it above is ot uique, we do kow that for ay ω Ω the fuctio t B t (ω) is cotiuous almost surely. Thus, every ω Ω gives rise to a cotiuous fuctio from [0, ) to R d. I this way, we ca thik of Browia motio as a radomly chose elemet i the set of cotiuous fuctios from [0, ) to R d accordig to the probability measure P. Or, i our cotext: a radom cotiuous path startig at the origi. ) Refereces [] Probability with Martigales, D. Williams, Cambridge Uiversity Press, 99. [2] Stochastic Differetial Equatios, B. Øksedal, Spriger-Verlag, 985. [3] Pricipals of Radom Walk, F. Spitzer, Spriger-Verlag, secod editio,

22 [4] Radom Walks ad Radom Eviromets, B. Hughes, Oxford Uiversity Press, 995. [5] Probability: Theory ad Examples, R. Durret, Duxbury Press, secod editio,

Convergence of random variables. (telegram style notes) P.J.C. Spreij

Convergence of random variables. (telegram style notes) P.J.C. Spreij Covergece of radom variables (telegram style otes).j.c. Spreij this versio: September 6, 2005 Itroductio As we kow, radom variables are by defiitio measurable fuctios o some uderlyig measurable space

More information

Infinite Sequences and Series

Infinite Sequences and Series Chapter 6 Ifiite Sequeces ad Series 6.1 Ifiite Sequeces 6.1.1 Elemetary Cocepts Simply speakig, a sequece is a ordered list of umbers writte: {a 1, a 2, a 3,...a, a +1,...} where the elemets a i represet

More information

6.3 Testing Series With Positive Terms

6.3 Testing Series With Positive Terms 6.3. TESTING SERIES WITH POSITIVE TERMS 307 6.3 Testig Series With Positive Terms 6.3. Review of what is kow up to ow I theory, testig a series a i for covergece amouts to fidig the i= sequece of partial

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 21 11/27/2013

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 21 11/27/2013 MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 21 11/27/2013 Fuctioal Law of Large Numbers. Costructio of the Wieer Measure Cotet. 1. Additioal techical results o weak covergece

More information

7.1 Convergence of sequences of random variables

7.1 Convergence of sequences of random variables Chapter 7 Limit Theorems Throughout this sectio we will assume a probability space (, F, P), i which is defied a ifiite sequece of radom variables (X ) ad a radom variable X. The fact that for every ifiite

More information

Chapter 3. Strong convergence. 3.1 Definition of almost sure convergence

Chapter 3. Strong convergence. 3.1 Definition of almost sure convergence Chapter 3 Strog covergece As poited out i the Chapter 2, there are multiple ways to defie the otio of covergece of a sequece of radom variables. That chapter defied covergece i probability, covergece i

More information

Random Models. Tusheng Zhang. February 14, 2013

Random Models. Tusheng Zhang. February 14, 2013 Radom Models Tusheg Zhag February 14, 013 1 Radom Walks Let me describe the model. Radom walks are used to describe the motio of a movig particle (object). Suppose that a particle (object) moves alog the

More information

4. Partial Sums and the Central Limit Theorem

4. Partial Sums and the Central Limit Theorem 1 of 10 7/16/2009 6:05 AM Virtual Laboratories > 6. Radom Samples > 1 2 3 4 5 6 7 4. Partial Sums ad the Cetral Limit Theorem The cetral limit theorem ad the law of large umbers are the two fudametal theorems

More information

Chapter 6 Infinite Series

Chapter 6 Infinite Series Chapter 6 Ifiite Series I the previous chapter we cosidered itegrals which were improper i the sese that the iterval of itegratio was ubouded. I this chapter we are goig to discuss a topic which is somewhat

More information

Advanced Stochastic Processes.

Advanced Stochastic Processes. Advaced Stochastic Processes. David Gamarik LECTURE 2 Radom variables ad measurable fuctios. Strog Law of Large Numbers (SLLN). Scary stuff cotiued... Outlie of Lecture Radom variables ad measurable fuctios.

More information

7.1 Convergence of sequences of random variables

7.1 Convergence of sequences of random variables Chapter 7 Limit theorems Throughout this sectio we will assume a probability space (Ω, F, P), i which is defied a ifiite sequece of radom variables (X ) ad a radom variable X. The fact that for every ifiite

More information

An Introduction to Randomized Algorithms

An Introduction to Randomized Algorithms A Itroductio to Radomized Algorithms The focus of this lecture is to study a radomized algorithm for quick sort, aalyze it usig probabilistic recurrece relatios, ad also provide more geeral tools for aalysis

More information

Lecture 4. We also define the set of possible values for the random walk as the set of all x R d such that P(S n = x) > 0 for some n.

Lecture 4. We also define the set of possible values for the random walk as the set of all x R d such that P(S n = x) > 0 for some n. Radom Walks ad Browia Motio Tel Aviv Uiversity Sprig 20 Lecture date: Mar 2, 20 Lecture 4 Istructor: Ro Peled Scribe: Lira Rotem This lecture deals primarily with recurrece for geeral radom walks. We preset

More information

Lecture 3 : Random variables and their distributions

Lecture 3 : Random variables and their distributions Lecture 3 : Radom variables ad their distributios 3.1 Radom variables Let (Ω, F) ad (S, S) be two measurable spaces. A map X : Ω S is measurable or a radom variable (deoted r.v.) if X 1 (A) {ω : X(ω) A}

More information

Measure and Measurable Functions

Measure and Measurable Functions 3 Measure ad Measurable Fuctios 3.1 Measure o a Arbitrary σ-algebra Recall from Chapter 2 that the set M of all Lebesgue measurable sets has the followig properties: R M, E M implies E c M, E M for N implies

More information

Distribution of Random Samples & Limit theorems

Distribution of Random Samples & Limit theorems STAT/MATH 395 A - PROBABILITY II UW Witer Quarter 2017 Néhémy Lim Distributio of Radom Samples & Limit theorems 1 Distributio of i.i.d. Samples Motivatig example. Assume that the goal of a study is to

More information

Sequences and Series of Functions

Sequences and Series of Functions Chapter 6 Sequeces ad Series of Fuctios 6.1. Covergece of a Sequece of Fuctios Poitwise Covergece. Defiitio 6.1. Let, for each N, fuctio f : A R be defied. If, for each x A, the sequece (f (x)) coverges

More information

Axioms of Measure Theory

Axioms of Measure Theory MATH 532 Axioms of Measure Theory Dr. Neal, WKU I. The Space Throughout the course, we shall let X deote a geeric o-empty set. I geeral, we shall ot assume that ay algebraic structure exists o X so that

More information

Math 155 (Lecture 3)

Math 155 (Lecture 3) Math 55 (Lecture 3) September 8, I this lecture, we ll cosider the aswer to oe of the most basic coutig problems i combiatorics Questio How may ways are there to choose a -elemet subset of the set {,,,

More information

Singular Continuous Measures by Michael Pejic 5/14/10

Singular Continuous Measures by Michael Pejic 5/14/10 Sigular Cotiuous Measures by Michael Peic 5/4/0 Prelimiaries Give a set X, a σ-algebra o X is a collectio of subsets of X that cotais X ad ad is closed uder complemetatio ad coutable uios hece, coutable

More information

Introduction to Probability. Ariel Yadin

Introduction to Probability. Ariel Yadin Itroductio to robability Ariel Yadi Lecture 2 *** Ja. 7 ***. Covergece of Radom Variables As i the case of sequeces of umbers, we would like to talk about covergece of radom variables. There are may ways

More information

n outcome is (+1,+1, 1,..., 1). Let the r.v. X denote our position (relative to our starting point 0) after n moves. Thus X = X 1 + X 2 + +X n,

n outcome is (+1,+1, 1,..., 1). Let the r.v. X denote our position (relative to our starting point 0) after n moves. Thus X = X 1 + X 2 + +X n, CS 70 Discrete Mathematics for CS Sprig 2008 David Wager Note 9 Variace Questio: At each time step, I flip a fair coi. If it comes up Heads, I walk oe step to the right; if it comes up Tails, I walk oe

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 6 9/23/2013. Brownian motion. Introduction

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 6 9/23/2013. Brownian motion. Introduction MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/5.070J Fall 203 Lecture 6 9/23/203 Browia motio. Itroductio Cotet.. A heuristic costructio of a Browia motio from a radom walk. 2. Defiitio ad basic properties

More information

Product measures, Tonelli s and Fubini s theorems For use in MAT3400/4400, autumn 2014 Nadia S. Larsen. Version of 13 October 2014.

Product measures, Tonelli s and Fubini s theorems For use in MAT3400/4400, autumn 2014 Nadia S. Larsen. Version of 13 October 2014. Product measures, Toelli s ad Fubii s theorems For use i MAT3400/4400, autum 2014 Nadia S. Larse Versio of 13 October 2014. 1. Costructio of the product measure The purpose of these otes is to preset the

More information

Fall 2013 MTH431/531 Real analysis Section Notes

Fall 2013 MTH431/531 Real analysis Section Notes Fall 013 MTH431/531 Real aalysis Sectio 8.1-8. Notes Yi Su 013.11.1 1. Defiitio of uiform covergece. We look at a sequece of fuctios f (x) ad study the coverget property. Notice we have two parameters

More information

Problem Set 2 Solutions

Problem Set 2 Solutions CS271 Radomess & Computatio, Sprig 2018 Problem Set 2 Solutios Poit totals are i the margi; the maximum total umber of poits was 52. 1. Probabilistic method for domiatig sets 6pts Pick a radom subset S

More information

Solution. 1 Solutions of Homework 1. Sangchul Lee. October 27, Problem 1.1

Solution. 1 Solutions of Homework 1. Sangchul Lee. October 27, Problem 1.1 Solutio Sagchul Lee October 7, 017 1 Solutios of Homework 1 Problem 1.1 Let Ω,F,P) be a probability space. Show that if {A : N} F such that A := lim A exists, the PA) = lim PA ). Proof. Usig the cotiuity

More information

Discrete Mathematics for CS Spring 2007 Luca Trevisan Lecture 22

Discrete Mathematics for CS Spring 2007 Luca Trevisan Lecture 22 CS 70 Discrete Mathematics for CS Sprig 2007 Luca Trevisa Lecture 22 Aother Importat Distributio The Geometric Distributio Questio: A biased coi with Heads probability p is tossed repeatedly util the first

More information

Definition 4.2. (a) A sequence {x n } in a Banach space X is a basis for X if. unique scalars a n (x) such that x = n. a n (x) x n. (4.

Definition 4.2. (a) A sequence {x n } in a Banach space X is a basis for X if. unique scalars a n (x) such that x = n. a n (x) x n. (4. 4. BASES I BAACH SPACES 39 4. BASES I BAACH SPACES Sice a Baach space X is a vector space, it must possess a Hamel, or vector space, basis, i.e., a subset {x γ } γ Γ whose fiite liear spa is all of X ad

More information

Notes 5 : More on the a.s. convergence of sums

Notes 5 : More on the a.s. convergence of sums Notes 5 : More o the a.s. covergece of sums Math 733-734: Theory of Probability Lecturer: Sebastie Roch Refereces: Dur0, Sectios.5; Wil9, Sectio 4.7, Shi96, Sectio IV.4, Dur0, Sectio.. Radom series. Three-series

More information

A sequence of numbers is a function whose domain is the positive integers. We can see that the sequence

A sequence of numbers is a function whose domain is the positive integers. We can see that the sequence Sequeces A sequece of umbers is a fuctio whose domai is the positive itegers. We ca see that the sequece,, 2, 2, 3, 3,... is a fuctio from the positive itegers whe we write the first sequece elemet as

More information

Ma 530 Infinite Series I

Ma 530 Infinite Series I Ma 50 Ifiite Series I Please ote that i additio to the material below this lecture icorporated material from the Visual Calculus web site. The material o sequeces is at Visual Sequeces. (To use this li

More information

Discrete Mathematics for CS Spring 2008 David Wagner Note 22

Discrete Mathematics for CS Spring 2008 David Wagner Note 22 CS 70 Discrete Mathematics for CS Sprig 2008 David Wager Note 22 I.I.D. Radom Variables Estimatig the bias of a coi Questio: We wat to estimate the proportio p of Democrats i the US populatio, by takig

More information

lim za n n = z lim a n n.

lim za n n = z lim a n n. Lecture 6 Sequeces ad Series Defiitio 1 By a sequece i a set A, we mea a mappig f : N A. It is customary to deote a sequece f by {s } where, s := f(). A sequece {z } of (complex) umbers is said to be coverget

More information

This exam contains 19 pages (including this cover page) and 10 questions. A Formulae sheet is provided with the exam.

This exam contains 19 pages (including this cover page) and 10 questions. A Formulae sheet is provided with the exam. Probability ad Statistics FS 07 Secod Sessio Exam 09.0.08 Time Limit: 80 Miutes Name: Studet ID: This exam cotais 9 pages (icludig this cover page) ad 0 questios. A Formulae sheet is provided with the

More information

Econ 325/327 Notes on Sample Mean, Sample Proportion, Central Limit Theorem, Chi-square Distribution, Student s t distribution 1.

Econ 325/327 Notes on Sample Mean, Sample Proportion, Central Limit Theorem, Chi-square Distribution, Student s t distribution 1. Eco 325/327 Notes o Sample Mea, Sample Proportio, Cetral Limit Theorem, Chi-square Distributio, Studet s t distributio 1 Sample Mea By Hiro Kasahara We cosider a radom sample from a populatio. Defiitio

More information

Random Variables, Sampling and Estimation

Random Variables, Sampling and Estimation Chapter 1 Radom Variables, Samplig ad Estimatio 1.1 Itroductio This chapter will cover the most importat basic statistical theory you eed i order to uderstad the ecoometric material that will be comig

More information

Sequences A sequence of numbers is a function whose domain is the positive integers. We can see that the sequence

Sequences A sequence of numbers is a function whose domain is the positive integers. We can see that the sequence Sequeces A sequece of umbers is a fuctio whose domai is the positive itegers. We ca see that the sequece 1, 1, 2, 2, 3, 3,... is a fuctio from the positive itegers whe we write the first sequece elemet

More information

Integrable Functions. { f n } is called a determining sequence for f. If f is integrable with respect to, then f d does exist as a finite real number

Integrable Functions. { f n } is called a determining sequence for f. If f is integrable with respect to, then f d does exist as a finite real number MATH 532 Itegrable Fuctios Dr. Neal, WKU We ow shall defie what it meas for a measurable fuctio to be itegrable, show that all itegral properties of simple fuctios still hold, ad the give some coditios

More information

Notes 27 : Brownian motion: path properties

Notes 27 : Brownian motion: path properties Notes 27 : Browia motio: path properties Math 733-734: Theory of Probability Lecturer: Sebastie Roch Refereces:[Dur10, Sectio 8.1], [MP10, Sectio 1.1, 1.2, 1.3]. Recall: DEF 27.1 (Covariace) Let X = (X

More information

If a subset E of R contains no open interval, is it of zero measure? For instance, is the set of irrationals in [0, 1] is of measure zero?

If a subset E of R contains no open interval, is it of zero measure? For instance, is the set of irrationals in [0, 1] is of measure zero? 2 Lebesgue Measure I Chapter 1 we defied the cocept of a set of measure zero, ad we have observed that every coutable set is of measure zero. Here are some atural questios: If a subset E of R cotais a

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 19 11/17/2008 LAWS OF LARGE NUMBERS II THE STRONG LAW OF LARGE NUMBERS

MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 19 11/17/2008 LAWS OF LARGE NUMBERS II THE STRONG LAW OF LARGE NUMBERS MASSACHUSTTS INSTITUT OF TCHNOLOGY 6.436J/5.085J Fall 2008 Lecture 9 /7/2008 LAWS OF LARG NUMBRS II Cotets. The strog law of large umbers 2. The Cheroff boud TH STRONG LAW OF LARG NUMBRS While the weak

More information

TR/46 OCTOBER THE ZEROS OF PARTIAL SUMS OF A MACLAURIN EXPANSION A. TALBOT

TR/46 OCTOBER THE ZEROS OF PARTIAL SUMS OF A MACLAURIN EXPANSION A. TALBOT TR/46 OCTOBER 974 THE ZEROS OF PARTIAL SUMS OF A MACLAURIN EXPANSION by A. TALBOT .. Itroductio. A problem i approximatio theory o which I have recetly worked [] required for its solutio a proof that the

More information

MAT1026 Calculus II Basic Convergence Tests for Series

MAT1026 Calculus II Basic Convergence Tests for Series MAT026 Calculus II Basic Covergece Tests for Series Egi MERMUT 202.03.08 Dokuz Eylül Uiversity Faculty of Sciece Departmet of Mathematics İzmir/TURKEY Cotets Mootoe Covergece Theorem 2 2 Series of Real

More information

Discrete Mathematics for CS Spring 2005 Clancy/Wagner Notes 21. Some Important Distributions

Discrete Mathematics for CS Spring 2005 Clancy/Wagner Notes 21. Some Important Distributions CS 70 Discrete Mathematics for CS Sprig 2005 Clacy/Wager Notes 21 Some Importat Distributios Questio: A biased coi with Heads probability p is tossed repeatedly util the first Head appears. What is the

More information

Problem Cosider the curve give parametrically as x = si t ad y = + cos t for» t» ß: (a) Describe the path this traverses: Where does it start (whe t =

Problem Cosider the curve give parametrically as x = si t ad y = + cos t for» t» ß: (a) Describe the path this traverses: Where does it start (whe t = Mathematics Summer Wilso Fial Exam August 8, ANSWERS Problem 1 (a) Fid the solutio to y +x y = e x x that satisfies y() = 5 : This is already i the form we used for a first order liear differetial equatio,

More information

On Random Line Segments in the Unit Square

On Random Line Segments in the Unit Square O Radom Lie Segmets i the Uit Square Thomas A. Courtade Departmet of Electrical Egieerig Uiversity of Califoria Los Ageles, Califoria 90095 Email: tacourta@ee.ucla.edu I. INTRODUCTION Let Q = [0, 1] [0,

More information

January 25, 2017 INTRODUCTION TO MATHEMATICAL STATISTICS

January 25, 2017 INTRODUCTION TO MATHEMATICAL STATISTICS Jauary 25, 207 INTRODUCTION TO MATHEMATICAL STATISTICS Abstract. A basic itroductio to statistics assumig kowledge of probability theory.. Probability I a typical udergraduate problem i probability, we

More information

Seunghee Ye Ma 8: Week 5 Oct 28

Seunghee Ye Ma 8: Week 5 Oct 28 Week 5 Summary I Sectio, we go over the Mea Value Theorem ad its applicatios. I Sectio 2, we will recap what we have covered so far this term. Topics Page Mea Value Theorem. Applicatios of the Mea Value

More information

MA131 - Analysis 1. Workbook 3 Sequences II

MA131 - Analysis 1. Workbook 3 Sequences II MA3 - Aalysis Workbook 3 Sequeces II Autum 2004 Cotets 2.8 Coverget Sequeces........................ 2.9 Algebra of Limits......................... 2 2.0 Further Useful Results........................

More information

(A sequence also can be thought of as the list of function values attained for a function f :ℵ X, where f (n) = x n for n 1.) x 1 x N +k x N +4 x 3

(A sequence also can be thought of as the list of function values attained for a function f :ℵ X, where f (n) = x n for n 1.) x 1 x N +k x N +4 x 3 MATH 337 Sequeces Dr. Neal, WKU Let X be a metric space with distace fuctio d. We shall defie the geeral cocept of sequece ad limit i a metric space, the apply the results i particular to some special

More information

6 Integers Modulo n. integer k can be written as k = qn + r, with q,r, 0 r b. So any integer.

6 Integers Modulo n. integer k can be written as k = qn + r, with q,r, 0 r b. So any integer. 6 Itegers Modulo I Example 2.3(e), we have defied the cogruece of two itegers a,b with respect to a modulus. Let us recall that a b (mod ) meas a b. We have proved that cogruece is a equivalece relatio

More information

Notes 19 : Martingale CLT

Notes 19 : Martingale CLT Notes 9 : Martigale CLT Math 733-734: Theory of Probability Lecturer: Sebastie Roch Refereces: [Bil95, Chapter 35], [Roc, Chapter 3]. Sice we have ot ecoutered weak covergece i some time, we first recall

More information

Lecture 19: Convergence

Lecture 19: Convergence Lecture 19: Covergece Asymptotic approach I statistical aalysis or iferece, a key to the success of fidig a good procedure is beig able to fid some momets ad/or distributios of various statistics. I may

More information

Lecture Notes for Analysis Class

Lecture Notes for Analysis Class Lecture Notes for Aalysis Class Topological Spaces A topology for a set X is a collectio T of subsets of X such that: (a) X ad the empty set are i T (b) Uios of elemets of T are i T (c) Fiite itersectios

More information

It is always the case that unions, intersections, complements, and set differences are preserved by the inverse image of a function.

It is always the case that unions, intersections, complements, and set differences are preserved by the inverse image of a function. MATH 532 Measurable Fuctios Dr. Neal, WKU Throughout, let ( X, F, µ) be a measure space ad let (!, F, P ) deote the special case of a probability space. We shall ow begi to study real-valued fuctios defied

More information

Entropy Rates and Asymptotic Equipartition

Entropy Rates and Asymptotic Equipartition Chapter 29 Etropy Rates ad Asymptotic Equipartitio Sectio 29. itroduces the etropy rate the asymptotic etropy per time-step of a stochastic process ad shows that it is well-defied; ad similarly for iformatio,

More information

Math 341 Lecture #31 6.5: Power Series

Math 341 Lecture #31 6.5: Power Series Math 341 Lecture #31 6.5: Power Series We ow tur our attetio to a particular kid of series of fuctios, amely, power series, f(x = a x = a 0 + a 1 x + a 2 x 2 + where a R for all N. I terms of a series

More information

STAT Homework 1 - Solutions

STAT Homework 1 - Solutions STAT-36700 Homework 1 - Solutios Fall 018 September 11, 018 This cotais solutios for Homework 1. Please ote that we have icluded several additioal commets ad approaches to the problems to give you better

More information

CS284A: Representations and Algorithms in Molecular Biology

CS284A: Representations and Algorithms in Molecular Biology CS284A: Represetatios ad Algorithms i Molecular Biology Scribe Notes o Lectures 3 & 4: Motif Discovery via Eumeratio & Motif Represetatio Usig Positio Weight Matrix Joshua Gervi Based o presetatios by

More information

A Proof of Birkhoff s Ergodic Theorem

A Proof of Birkhoff s Ergodic Theorem A Proof of Birkhoff s Ergodic Theorem Joseph Hora September 2, 205 Itroductio I Fall 203, I was learig the basics of ergodic theory, ad I came across this theorem. Oe of my supervisors, Athoy Quas, showed

More information

HOMEWORK 2 SOLUTIONS

HOMEWORK 2 SOLUTIONS HOMEWORK SOLUTIONS CSE 55 RANDOMIZED AND APPROXIMATION ALGORITHMS 1. Questio 1. a) The larger the value of k is, the smaller the expected umber of days util we get all the coupos we eed. I fact if = k

More information

Generalized Semi- Markov Processes (GSMP)

Generalized Semi- Markov Processes (GSMP) Geeralized Semi- Markov Processes (GSMP) Summary Some Defiitios Markov ad Semi-Markov Processes The Poisso Process Properties of the Poisso Process Iterarrival times Memoryless property ad the residual

More information

Randomized Algorithms I, Spring 2018, Department of Computer Science, University of Helsinki Homework 1: Solutions (Discussed January 25, 2018)

Randomized Algorithms I, Spring 2018, Department of Computer Science, University of Helsinki Homework 1: Solutions (Discussed January 25, 2018) Radomized Algorithms I, Sprig 08, Departmet of Computer Sciece, Uiversity of Helsiki Homework : Solutios Discussed Jauary 5, 08). Exercise.: Cosider the followig balls-ad-bi game. We start with oe black

More information

Mathematical Methods for Physics and Engineering

Mathematical Methods for Physics and Engineering Mathematical Methods for Physics ad Egieerig Lecture otes Sergei V. Shabaov Departmet of Mathematics, Uiversity of Florida, Gaiesville, FL 326 USA CHAPTER The theory of covergece. Numerical sequeces..

More information

Lecture Chapter 6: Convergence of Random Sequences

Lecture Chapter 6: Convergence of Random Sequences ECE5: Aalysis of Radom Sigals Fall 6 Lecture Chapter 6: Covergece of Radom Sequeces Dr Salim El Rouayheb Scribe: Abhay Ashutosh Doel, Qibo Zhag, Peiwe Tia, Pegzhe Wag, Lu Liu Radom sequece Defiitio A ifiite

More information

Lecture 3 The Lebesgue Integral

Lecture 3 The Lebesgue Integral Lecture 3: The Lebesgue Itegral 1 of 14 Course: Theory of Probability I Term: Fall 2013 Istructor: Gorda Zitkovic Lecture 3 The Lebesgue Itegral The costructio of the itegral Uless expressly specified

More information

Math 113, Calculus II Winter 2007 Final Exam Solutions

Math 113, Calculus II Winter 2007 Final Exam Solutions Math, Calculus II Witer 7 Fial Exam Solutios (5 poits) Use the limit defiitio of the defiite itegral ad the sum formulas to compute x x + dx The check your aswer usig the Evaluatio Theorem Solutio: I this

More information

Lecture 6: Integration and the Mean Value Theorem. slope =

Lecture 6: Integration and the Mean Value Theorem. slope = Math 8 Istructor: Padraic Bartlett Lecture 6: Itegratio ad the Mea Value Theorem Week 6 Caltech 202 The Mea Value Theorem The Mea Value Theorem abbreviated MVT is the followig result: Theorem. Suppose

More information

sin(n) + 2 cos(2n) n 3/2 3 sin(n) 2cos(2n) n 3/2 a n =

sin(n) + 2 cos(2n) n 3/2 3 sin(n) 2cos(2n) n 3/2 a n = 60. Ratio ad root tests 60.1. Absolutely coverget series. Defiitio 13. (Absolute covergece) A series a is called absolutely coverget if the series of absolute values a is coverget. The absolute covergece

More information

Lecture 2: Concentration Bounds

Lecture 2: Concentration Bounds CSE 52: Desig ad Aalysis of Algorithms I Sprig 206 Lecture 2: Cocetratio Bouds Lecturer: Shaya Oveis Ghara March 30th Scribe: Syuzaa Sargsya Disclaimer: These otes have ot bee subjected to the usual scrutiy

More information

University of Colorado Denver Dept. Math. & Stat. Sciences Applied Analysis Preliminary Exam 13 January 2012, 10:00 am 2:00 pm. Good luck!

University of Colorado Denver Dept. Math. & Stat. Sciences Applied Analysis Preliminary Exam 13 January 2012, 10:00 am 2:00 pm. Good luck! Uiversity of Colorado Dever Dept. Math. & Stat. Scieces Applied Aalysis Prelimiary Exam 13 Jauary 01, 10:00 am :00 pm Name: The proctor will let you read the followig coditios before the exam begis, ad

More information

Math 216A Notes, Week 5

Math 216A Notes, Week 5 Math 6A Notes, Week 5 Scribe: Ayastassia Sebolt Disclaimer: These otes are ot early as polished (ad quite possibly ot early as correct) as a published paper. Please use them at your ow risk.. Thresholds

More information

Probability, Expectation Value and Uncertainty

Probability, Expectation Value and Uncertainty Chapter 1 Probability, Expectatio Value ad Ucertaity We have see that the physically observable properties of a quatum system are represeted by Hermitea operators (also referred to as observables ) such

More information

Chapter 6 Principles of Data Reduction

Chapter 6 Principles of Data Reduction Chapter 6 for BST 695: Special Topics i Statistical Theory. Kui Zhag, 0 Chapter 6 Priciples of Data Reductio Sectio 6. Itroductio Goal: To summarize or reduce the data X, X,, X to get iformatio about a

More information

x a x a Lecture 2 Series (See Chapter 1 in Boas)

x a x a Lecture 2 Series (See Chapter 1 in Boas) Lecture Series (See Chapter i Boas) A basic ad very powerful (if pedestria, recall we are lazy AD smart) way to solve ay differetial (or itegral) equatio is via a series expasio of the correspodig solutio

More information

Section 1.1. Calculus: Areas And Tangents. Difference Equations to Differential Equations

Section 1.1. Calculus: Areas And Tangents. Difference Equations to Differential Equations Differece Equatios to Differetial Equatios Sectio. Calculus: Areas Ad Tagets The study of calculus begis with questios about chage. What happes to the velocity of a swigig pedulum as its positio chages?

More information

Chapter 0. Review of set theory. 0.1 Sets

Chapter 0. Review of set theory. 0.1 Sets Chapter 0 Review of set theory Set theory plays a cetral role i the theory of probability. Thus, we will ope this course with a quick review of those otios of set theory which will be used repeatedly.

More information

Dirichlet s Theorem on Arithmetic Progressions

Dirichlet s Theorem on Arithmetic Progressions Dirichlet s Theorem o Arithmetic Progressios Athoy Várilly Harvard Uiversity, Cambridge, MA 0238 Itroductio Dirichlet s theorem o arithmetic progressios is a gem of umber theory. A great part of its beauty

More information

5 Birkhoff s Ergodic Theorem

5 Birkhoff s Ergodic Theorem 5 Birkhoff s Ergodic Theorem Amog the most useful of the various geeralizatios of KolmogorovâĂŹs strog law of large umbers are the ergodic theorems of Birkhoff ad Kigma, which exted the validity of the

More information

Discrete Mathematics and Probability Theory Summer 2014 James Cook Note 15

Discrete Mathematics and Probability Theory Summer 2014 James Cook Note 15 CS 70 Discrete Mathematics ad Probability Theory Summer 2014 James Cook Note 15 Some Importat Distributios I this ote we will itroduce three importat probability distributios that are widely used to model

More information

Basics of Probability Theory (for Theory of Computation courses)

Basics of Probability Theory (for Theory of Computation courses) Basics of Probability Theory (for Theory of Computatio courses) Oded Goldreich Departmet of Computer Sciece Weizma Istitute of Sciece Rehovot, Israel. oded.goldreich@weizma.ac.il November 24, 2008 Preface.

More information

Discrete Mathematics and Probability Theory Spring 2016 Rao and Walrand Note 19

Discrete Mathematics and Probability Theory Spring 2016 Rao and Walrand Note 19 CS 70 Discrete Mathematics ad Probability Theory Sprig 2016 Rao ad Walrad Note 19 Some Importat Distributios Recall our basic probabilistic experimet of tossig a biased coi times. This is a very simple

More information

Rates of Convergence by Moduli of Continuity

Rates of Convergence by Moduli of Continuity Rates of Covergece by Moduli of Cotiuity Joh Duchi: Notes for Statistics 300b March, 017 1 Itroductio I this ote, we give a presetatio showig the importace, ad relatioship betwee, the modulis of cotiuity

More information

Sequences. Notation. Convergence of a Sequence

Sequences. Notation. Convergence of a Sequence Sequeces A sequece is essetially just a list. Defiitio (Sequece of Real Numbers). A sequece of real umbers is a fuctio Z (, ) R for some real umber. Do t let the descriptio of the domai cofuse you; it

More information

ACO Comprehensive Exam 9 October 2007 Student code A. 1. Graph Theory

ACO Comprehensive Exam 9 October 2007 Student code A. 1. Graph Theory 1. Graph Theory Prove that there exist o simple plaar triagulatio T ad two distict adjacet vertices x, y V (T ) such that x ad y are the oly vertices of T of odd degree. Do ot use the Four-Color Theorem.

More information

Probability for mathematicians INDEPENDENCE TAU

Probability for mathematicians INDEPENDENCE TAU Probability for mathematicias INDEPENDENCE TAU 2013 28 Cotets 3 Ifiite idepedet sequeces 28 3a Idepedet evets........................ 28 3b Idepedet radom variables.................. 33 3 Ifiite idepedet

More information

Solutions to HW Assignment 1

Solutions to HW Assignment 1 Solutios to HW: 1 Course: Theory of Probability II Page: 1 of 6 Uiversity of Texas at Austi Solutios to HW Assigmet 1 Problem 1.1. Let Ω, F, {F } 0, P) be a filtered probability space ad T a stoppig time.

More information

Problems from 9th edition of Probability and Statistical Inference by Hogg, Tanis and Zimmerman:

Problems from 9th edition of Probability and Statistical Inference by Hogg, Tanis and Zimmerman: Math 224 Fall 2017 Homework 4 Drew Armstrog Problems from 9th editio of Probability ad Statistical Iferece by Hogg, Tais ad Zimmerma: Sectio 2.3, Exercises 16(a,d),18. Sectio 2.4, Exercises 13, 14. Sectio

More information

(b) What is the probability that a particle reaches the upper boundary n before the lower boundary m?

(b) What is the probability that a particle reaches the upper boundary n before the lower boundary m? MATH 529 The Boudary Problem The drukard s walk (or boudary problem) is oe of the most famous problems i the theory of radom walks. Oe versio of the problem is described as follows: Suppose a particle

More information

62. Power series Definition 16. (Power series) Given a sequence {c n }, the series. c n x n = c 0 + c 1 x + c 2 x 2 + c 3 x 3 +

62. Power series Definition 16. (Power series) Given a sequence {c n }, the series. c n x n = c 0 + c 1 x + c 2 x 2 + c 3 x 3 + 62. Power series Defiitio 16. (Power series) Give a sequece {c }, the series c x = c 0 + c 1 x + c 2 x 2 + c 3 x 3 + is called a power series i the variable x. The umbers c are called the coefficiets of

More information

1 The Haar functions and the Brownian motion

1 The Haar functions and the Brownian motion 1 The Haar fuctios ad the Browia motio 1.1 The Haar fuctios ad their completeess The Haar fuctios The basic Haar fuctio is 1 if x < 1/2, ψx) = 1 if 1/2 x < 1, otherwise. 1.1) It has mea zero 1 ψx)dx =,

More information

The picture in figure 1.1 helps us to see that the area represents the distance traveled. Figure 1: Area represents distance travelled

The picture in figure 1.1 helps us to see that the area represents the distance traveled. Figure 1: Area represents distance travelled 1 Lecture : Area Area ad distace traveled Approximatig area by rectagles Summatio The area uder a parabola 1.1 Area ad distace Suppose we have the followig iformatio about the velocity of a particle, how

More information

The Borel hierarchy classifies subsets of the reals by their topological complexity. Another approach is to classify them by size.

The Borel hierarchy classifies subsets of the reals by their topological complexity. Another approach is to classify them by size. Lecture 7: Measure ad Category The Borel hierarchy classifies subsets of the reals by their topological complexity. Aother approach is to classify them by size. Filters ad Ideals The most commo measure

More information

Statistics 511 Additional Materials

Statistics 511 Additional Materials Cofidece Itervals o mu Statistics 511 Additioal Materials This topic officially moves us from probability to statistics. We begi to discuss makig ifereces about the populatio. Oe way to differetiate probability

More information

Lecture 10 October Minimaxity and least favorable prior sequences

Lecture 10 October Minimaxity and least favorable prior sequences STATS 300A: Theory of Statistics Fall 205 Lecture 0 October 22 Lecturer: Lester Mackey Scribe: Brya He, Rahul Makhijai Warig: These otes may cotai factual ad/or typographic errors. 0. Miimaxity ad least

More information

Introduction to Extreme Value Theory Laurens de Haan, ISM Japan, Erasmus University Rotterdam, NL University of Lisbon, PT

Introduction to Extreme Value Theory Laurens de Haan, ISM Japan, Erasmus University Rotterdam, NL University of Lisbon, PT Itroductio to Extreme Value Theory Laures de Haa, ISM Japa, 202 Itroductio to Extreme Value Theory Laures de Haa Erasmus Uiversity Rotterdam, NL Uiversity of Lisbo, PT Itroductio to Extreme Value Theory

More information

K. Grill Institut für Statistik und Wahrscheinlichkeitstheorie, TU Wien, Austria

K. Grill Institut für Statistik und Wahrscheinlichkeitstheorie, TU Wien, Austria MARKOV PROCESSES K. Grill Istitut für Statistik ud Wahrscheilichkeitstheorie, TU Wie, Austria Keywords: Markov process, Markov chai, Markov property, stoppig times, strog Markov property, trasitio matrix,

More information

1 Approximating Integrals using Taylor Polynomials

1 Approximating Integrals using Taylor Polynomials Seughee Ye Ma 8: Week 7 Nov Week 7 Summary This week, we will lear how we ca approximate itegrals usig Taylor series ad umerical methods. Topics Page Approximatig Itegrals usig Taylor Polyomials. Defiitios................................................

More information

Massachusetts Institute of Technology

Massachusetts Institute of Technology Solutios to Quiz : Sprig 006 Problem : Each of the followig statemets is either True or False. There will be o partial credit give for the True False questios, thus ay explaatios will ot be graded. Please

More information