The VC-dimension of Unions: Learning, Geometry and Combinatorics

Size: px
Start display at page:

Download "The VC-dimension of Unions: Learning, Geometry and Combinatorics"

Transcription

1 The VC-dimension of Unions: Leaning, Geomety and Combinatoics Mónika Csikós Andey Kupavskii Nabil H. Mustafa Abstact The VC-dimension of a set system is a way to captue its complexity, and has been a key paamete studied extensively in machine leaning and geomety communities. In this pape, we make substantial pogess on bounding the VC-dimension of k-fold unions and intesections of basic geometic set systems, including settling an open question in machine leaning that was fist studied in the 1989 foundational pape of Blume, Ehenfeucht, Haussle and Wamuth [4]. Kalsuhe Institute of Technology, Kalsuhe, Gemany. Moscow Institute of Physics and Technology and EPFL, Switzeland. kupavskii@yandex.u. The wok of Andey Kupavskii has been suppoted in pat by the Swiss National Science Foundation Gants and and by the gant N of the Russian Foundation fo Basic Reseach. Univesité Pais-Est, LIGM, Equipe A3SI, ESIEE Pais, Fance. mustafan@esiee.f. The wok of Nabil H. Mustafa has been suppoted by the gant ANR SAGA JCJC-14-CE ).

2 1 Intoduction Let X, R) be a set system, whee X is a set of elements and R is a set of subsets of X. In the theoy of leaning, each set of R is also called a concept, and R a concept class on X. Fo any intege k 2, define the k-fold union of R to be the set system R k = {R 1 R k : R 1,..., R k R}. Similaly, one can define the k-fold intesection of R as the set system R k consisting of all subsets deived fom the common intesection of at most k subsets of R. Note that as the subsets R 1,..., R k need not necessaily be distinct, we have R R k and R R k. VC-dimension. One of the fundamental measues fo captuing the complexity o ichness of a set system, with applications acoss seveal fields, is the Vapnik-Chevonenkis dimension, o in shot the VC-dimension, of a set system. Given X, R), fo any set Y X, define the pojection of R onto Y as the set system: R Y = {Y R: R R}. We say that R shattes Y if R Y = 2 Y ; in othe wods, all subsets of Y can be deived by intesection with sets of R. Then the VC-dimension of R, denoted VC-dimR), is the size of the lagest subset of X that can be shatteed by R. Oiginally defined in statistics and pobability, it has tuned out to be a key paamete in seveal aeas; this pape concens thee of them leaning theoy, geomety and combinatoics. Leaning theoy. In leaning theoy, the VC-dimension of a concept class measues the difficulty of leaning a concept of the class. The foundational pape of Blume, Ehenfeucht, Haussle and Wamuth [4] states that the essential condition fo distibution-fee leanability is finiteness of the Vapnik-Chevonenkis dimension. One of the theoems they pove is the following. Theoem A Blume et al. [4]). Let X, R) be a set system, and k be any positive intege. Then Ω VC-dim R) k) = VC-dim R k ) ) = O VC-dim R) k log k. Ω VC-dim R) k) = VC-dim R k ) ) = O VC-dim R) k log k. They also consideed the question of whethe the bound of Theoem A is tight in the most basic geometic case when X R d is a set of points and R is the pojection of the family of all half-spaces of R d onto X. Hee they poved that the VC dimension of the k-fold union of half-spaces in two dimensions is exactly 2k + 1. Fo geneal dimensions d 4, they bound the VC-dimension of the k-fold union of half-spaces by Od k log k), following fom Theoem A togethe with the fact that the VC-dimension of the set system induced by half-spaces in R d is d + 1. The same bound holds fo the k-fold intesection of half-spaces in R d. Eisenstat and Angluin [8] poved, by giving a pobabilistic constuction of an abstact set system, that the uppe bound of Theoem A is asymptotically tight if VC-dim R) 5 and that fo VC-dim R) = 1, a lowe bound of k holds and that it is tight. A few yeas late, Eisenstat [7] 1

3 filled the gap by showing that VC-dim R k ) = Ω VC-dim R) k log k) even if VC-dim R) 2. Late Dobkin and Gunopulos [6] showed that the VC-dimension of the k-fold union of half-spaces in thee dimensions is uppe-bounded by 4k. Fo d 4, the cuent best uppe-bound fo the k-fold union and the k-fold intesection of halfspaces in R d is still the one given by Theoem A almost 30 yeas ago, while the lowe-bound has emained Ω VC-dimR) k). We efe the eade to the PhD thesis [11] fo a summay of the bounds on VC-dimensions of these basic combinatoial and geometic set systems. The esolution of the VC-dimension of k-fold unions and intesections of half-spaces is left as one of the main open poblems in the thesis. In fact, while the uppe-bound of Theoem A applies to geometic set systems, we did not have, till now, a single example of a geometic set system R with a non-linea lowe-bound i.e., beyond Ω VC-dim R) k) on the VC-dimension of its k-fold union o k-fold intesection. Ou fist esult poves a non-linea bound fo a geometic set system even in the plane: Theoem 1 Section 3). Let k be a given positive intege. Then thee exists a set P of points in R 2 such that the set system L induced on P by lines in the plane satisfies VC-dim L k ) ) 1 ) ) 1 ) log k 3 log k 3 = Ω VC-dimL) k = Ω k. log log k log log k Next we esolve the question fo half-spaces in R d, d 4. Theoem 2 Section 4). Let k be a given positive intege, and d 4 an intege. Then thee exists a set P of points in R d such that the set system R induced on P by half-spaces satisfies VC-dim R k ) ) ) = Ω VC-dimR) k log k = Ω d k log k. VC-dim R k ) ) ) = Ω VC-dimR) k log k = Ω d k log k. Relations to Geomety. The beakthough wok of Haussle and Wezl [10] showed that the size of a key stuctue in discete and computational geomety, ɛ-nets we efe the eade to the books [5, Chapte 4], [13, Chapte 10], [16, Chapte 15] fo detailed infomation), is diectly linked with the VC-dimension of the coesponding set system: Theoem B Epsilon-net Theoem [10]). Let X, R) be a set system and ɛ > 0 be a given paamete. Then thee exists an ɛ-net fo X, R) of size at most VC-dimR) O log 1 ). ɛ ɛ Thus the sizes of ɛ-nets can be uppe-bounded in tems of the VC-dimension of the set system. On the othe hand, Alon [1] showed a supe-linea lowe-bound fo ɛ-nets induced by lines in the plane; in a vey ecent beakthough, Balogh and Solymosi [3] impoved Alon s lowe bound to get the following: 2

4 Theoem C [3]). Given any ɛ > 0, thee exists a set P of points in the plane such that any ɛ-net fo P fo the set system induced on P by lines must have size at least ) 1 1 log 1 3 ɛ 2ɛ log log 1. ɛ We efe the eade to the chapte [14] fo futhe details on ecent pogess in the aea of ɛ-nets. 2 Ou Results Ou esults exploit links between leaning theoy, geomety and combinatoics: the open poblem on the VC-dimension of the k-fold union of half-spaces in R d in leaning theoy is esolved by noting its connection to ɛ-nets in geomety items 1. and 2. below), and the ecent beakthough of Balogh and Solymosi on ɛ-nets can be genealized by putting it in the famewok of VC-dimension of k-fold unions item 3. below). Moe pecisely: 1. The stating point of ou wok is the following obsevation that links the study of k-fold unions in leaning theoy liteatue to the one fo ɛ-nets in discete geomety. While consideable wok has been done in the pecise elation between VC-dimension of set systems and uppe-bounds on the sizes of ɛ-nets, it tuns out the connection goes both ways: Theoem 3 Section 5). Let X, R) be a set system and ɛ > 0 be a given paamete. Then any ɛ-net fo X, R) must have size at least )) VC-dimR) 1 Ω f, ɛ 2ɛ whee f : R R is any function satisfying VC-dim R k ) VC-dimR) k fk) fo k = 1 2ɛ. This can be seen as complement to Theoem B: the VC-dimension of the k-fold union gives a lowe bound fo the sizes of ɛ-nets fo R. 2. We show an optimal lowe-bound on the VC-dimension of the k-fold union and the k-fold intesection of half-spaces in R d follows via [12], thus settling affimatively one of the main open question studied by Eisenstat and Angluin [8], Johnson [11], and Eisenstat [7], and matching the Od k log k) uppe bound of Theoem A. Theoem 2 Section 4). Let k be a given positive intege, and d 4 an intege. Then thee exists a set P of points in R d such that the set system R induced on P by half-spaces satisfies VC-dim R k ) ) ) = Ω VC-dimR) k log k = Ω d k log k. VC-dim R k ) ) ) = Ω VC-dimR) k log k = Ω d k log k. 3

5 Remak 1. This statement also povides a non-pobabilistic poof of Eisenstat s lowe-bound [7]. Remak 2. Obseve that if R := {R d \ R R R}, then VC-dimR) = VC-dimR) and so R k = R k. holds by the De Mogan laws. Since fo half-spaces R = R, the fist claim of Theoem 2 implies the second one, i.e., the same lowe-bound fo R k, settling anothe question posed by Eisenstat and Angluin [8]. Remak 3. This implies that the coeset constuctions in [9] equie an additional log k facto in the coeset size coming fom the VC-dimension of the k-fold intesection of half-spaces. See [2] fo details. 3. We next tun to the set system induced on a set of points in the plane by lines. We show that the poweful technique of Balogh and Solymosi [3] can be impoved and extended to pove a lowe-bound on the VC-dimension of the k-fold union of this set system induced by lines in the plane: Theoem 1 Section 3). Let k be a given positive intege. Then thee exists a set P of points in R 2 such that the set system L induced on P by lines in the plane satisfies VC-dim L k ) ) 1 ) ) 1 ) log k 3 log k 3 = Ω VC-dimL) k = Ω k. log log k log log k As an immediate coollay using Theoem 3), we get an impovement to Theoem C: Coollay 4. Given any ɛ > 0, thee exists a set P of points in the plane such that any ɛ-net fo P fo the set system induced on P by lines must have size ) 1 Ω 1 log 1 3 ɛ ɛ log log 1. ɛ 3 Poof of Theoem 1. We fist pesent the poof outline: we stat by taking an appopiate point set P in R fo a paamete to be chosen late), with the popety that P and all its sufficiently dense subsets contain many -tuples of collinea points. We will then pick each point of P independently with pobability p a paamete to be chosen late) to get the equied point-set R. The key idea is the use of the poweful containe method to analyze the failue pobability, which in ou case will be that R contains a subset of size m a paamete to be chosen late) without any collinea points. A andom pojection onto the plane then gives the equied point set. Ou poof closely follows the one of Balogh and Solymosi, with two key diffeences: i) one can choose the constuction and paametes moe caefully to impove the bound by poly log log factos. The initial set in [3] is simply the gid [n] ; we will pick a moe subtle constuction with fine popeties which allow us to get an impoved bound, and 4

6 ii) the use of containes is flexible and poweful enough to allow one to deduce the stonge statement on lowe-bound fo the VC dimension of the k-fold union. The value of m above in [3] is set to R 2, following the initial wok of Alon [1] stated as the Alon idea in [3]. This suffices to get a lowe-bound on ɛ-nets, but not fo lowe-bounding the VC-dimension of the k-fold union of the set system on R induced by lines. Towads this end, we need to pove a stonge popety with a lowe value of m. Now we tun to the poof of the theoem. 3.1 Initial set P : pointset with many collinea tuples Denote by [a, b] the set of all integes between a and b. In what follows, all asymptotic notation will be with espect to the paamete n; fo bevity, we will denote fn) = gn) ± ogn)) by fn) gn), and fn) gn) + ogn)) by fn) gn). Let n, N and t R be paametes to be fixed late, such that 1 t = on). Ou initial stating point-set in R will be: { } P := x 1,..., x ) [ n, n] : x i n x 1 fo i = 2,...,. Now conside the family L t of all lines that pass though at least one point of P, and have the diection vecto in the following set L t : { L t := x 1,..., x ): n t x 1 2n t, x i n } t fo i = 2,...,, and x 1,..., x ae elatively pime. Note that if x 1,..., x ) ae elatively pime integes, then αx 1,..., αx ) Z if and only if α Z. 1) We ae going to wok with the set of points P and the subsets of P induced by lines in L t. The key popeties of P and L t ae summaized in the following claim. Claim 4.1. Let P and L t be as defined above. Then i) P 2n), ii) L t 2n) 2t, iii) each point of P lies on L t lines of L t, iv) each line of L t contains at most 2t points fom P, and v) 2n) 2t L t L t 22n) t L t. Poof. i) P = 2n + 1) n i=1 2 n i) + 1) 1 2n). 5

7 ii) It follows fom Nymann [15] that the numbe of k-tuples of integes fom [m] x that ae elatively pime is asymptotically mx ζx), whee ζx) = i=1 i x is the Riemann zeta function. It is easy to see that ζx) 1 as x. Theefoe, even looking at x 2,..., x [ n t, n ] t, we conclude that the numbe of such 1)-tuples that ae elatively pime is asymptotically 2n t ) 1. Theefoe, almost all -tuples in [ n t, 2n ] [ t n t, n ] 1 t ae elatively pime, and the numbe of -tuples in L is asymptotically 2n) 2t. iii) Follows fom the definition of L t. iv) Popety 1) implies that if p P lies on a line with diection vecto v L t with coefficient α, then α must be an intege. The claim follows as the fist coodinate of p lies in the inteval [ n, n] while the absolute value of the fist coodinate of the diection vecto v is at least n t. v) Note that evey line in L t passes though one of the points of the following set: U t := { x 1,..., x ): x 1 2n } t, x i [ n, n] fo i = 2,...,. Theefoe we can uppe-bound the size of L t by the numbe of incidences with U t, i.e., L t L t U t 2 2n) t L t. On the othe hand, each line fom L t passes though at most fou points of U t fo the same easons as the ones used to pove iv). So each line is counted at most fou times in the numbe of incidences with the set U t, which gives the lowe bound of L t Lt Ut 4. Lemma 5. Let γ 82 t be a given positive paamete. Then evey subset of P of size γ P contains γ n 2 t 2 collinea -tuples lying on lines fom L t. Poof. Let S be any subset of P with S = γ P. Then Claim 4.1iii) implies that the numbe of point-line incidences between S and L t is S L t = γ P L t. Thus the aveage numbe of points of S lying on a line of L t, using Claim 4.1 v), is γ P L t L t γ P L t L t 22n) t γ 2n) 22n) t = γt 2. Note that by the choice of γ, we have γt 2 4, and thus γt 2 ) γt 4. By the convexity of ) x k and Claim 4.1v), we can conclude that the numbe of collinea -tuples in S is at least γt ) γt 2 2 L t )! 2n) L t 2t γt ) 4! 2n) L t 2t whee the last inequality used the bound 4! /2). γt ) 4 2n) 2n)! 2t 2t = γ n 2 1 t 4! γ n 2 t 2, 6

8 3.2 Setting up containes In this section we pepae notations fo, and state the main technical tool the containe theoem fo hypegaphs. Roughly speaking, containe theoems state that all independent sets in a hypegaph ae contained in a elatively small numbe of subsets, each of which is elatively small. Then, afte taking a andom subset R of vetices of the hypegaph, this type of statement helps us to bound the size of the lagest independent set in R by simply applying union bound ove all subsets of each of the containes. Unfotunately, containe statements ae athe technical. Let us fist intoduce some notation. Let H be a -unifom hypegaph on vetex set V H) and let d be the aveage degee of a vetex of H. The degee ds) of a subset S V H) is simply the numbe of edges of H containing S. Fo evey j [], let us denote by j the maximum degee of a j-element subset, that is, j := max {ds): S V H), S = j}. Fo any S V H), let H[S] denote the hypegaph induced by S, and by eh[s]) its numbe of hypeedges. Fo any τ 0, 1), define H, τ) := j=2 j d τ j 1 2 j 1 2 )+1 2). The ole of H, τ), oughly speaking, is to measue the goodness of H: smalle values of H, τ) imply bette bounds. Note that, fom this pespective, we ae inteested in lowe-bounding d and uppe-bounding j. We will use the following vaiant of a containe statement: Theoem D [18]). Let H be a -unifom hypegaph on the vetex set {1,..., N}. Let 0 < ε, τ < 1/2. Suppose that τ < 1/200! 2 ) and H, τ) ε/12!). Then thee exists c = c) 1000! 3 and a collection of vetex subsets C such that 1) evey independent set in H is a subset of some A C, 2) fo evey A C, eh[a]) ε eh), and 3) log C cnτ log 1 ε) log 1 τ ). We will wok with the -unifom hypegaph H, whee V H) := P, and edges ae all -tuples of collinea points on lines fom L t. Note that L t depends on a paamete t, which we set to be t = 5 2. Next we calculate the paametes of H that ae needed to apply Theoem D. Applying Lemma 5 with γ = 1 note hee that 82 t = 16 < 1), the total numbe of collinea -tuples in P is lowebounded by n2, and so the aveage degee d of H can be lowe-bounded by 3 t 2 d n 2 t 2 P = 2n n) = n n 3+2. On the othe hand, since each line contains at most 2t = 5 points, we get that 5 ) j j 5 j) fo j = 2,...,. j 7

9 Set τ = n and be such that 1 log n 20 log log n. Then note the following inequalities: n τ j 1 = n n j 1+ j = n j n j 2 > n j n 1 2 fo all j = 1,...,. j 2 2 j j) = n o j) fo all j = 2,..., 1. log n log n) 20 log log n = n Theefoe, we may conclude that fo this choice of τ and, we have H, τ) j=2 j=2 j d τ j 1 2 j 1 2 )+1 2) n o j) o1) n j n 1 2 n 1 3. j=2 5 j)+3+2 n τ j 1 2 j2 2 5 j) j 2 j=2 n τ j 1 We apply Theoem D with ou choice of τ and, and with ɛ = n 1 4. It is easy to see that the conditions of the theoem ae satisfied. Indeed, τ n 1 n 1 6 +o1) 2 1/200! 2 ), as well as that H, τ) n 1 3 = n ε ε 12!. Note that the constant c) fom the theoem satisfies c) 3. The conclusion of the theoem gives a collection of subsets C 2 P, such that fo evey A C we have eh[a]) n 1 4 eh) n 2 1 4, 2) whee the second inequality is due to the fact that eh) n 2, and such that log C 3 n n log 2 n 3 n ) Hee the last inequality is due to the fact that log 2 2 log log n n = n log n n Using Lemma 5 with γ = 2 which is a valid choice of γ, since γt = 3 2 ), we conclude that in any subset of size at least γ P thee ae at least n 2 / 4+o) n o1) induced hypeedges. Theefoe, any subset A C has size at most γ P by inequality 2). 3.3 Constucting R In this section we use the popeties of the hypegaph H, developed in Section 3.2, to obtain a point set fo which the VC-dimension of the k-fold union of the set system induced by lines is bounded fom below. 8

10 We would like to use the point set P with lines fom L t diectly, but the poblem is that it contains many collinea + 1)-tuples. Fom Claim 4.1 iv), the numbe of collinea + 1)-tuples in the set P with lines fom L t is at most 2t + 1 ) L t 2t)+1 + 1)! 22n) t 2n) 2t = 2t)+1 2n) 2 + 1)!t +1 = )! n2 24 ) n 2. Let R be a andom subset constucted by picking each point of P with pobability p, a paamete we will set shotly. We want the numbe of points in R to be much bigge than the numbe of collinea + 1)-tuples pesent in R. In othe wods, we set p such that E [ R ] = p P p 2n) p ) n 2. 4) Using Makov s inequality, it is easy to see that this inequality is satisfied, with high pobability, fo p = 20n. Set m = p P. Now we ague that, with high pobability, R has the popety that any subset of 3/2 R of size m is not an independent set; o stated anothe way, any subset of R of size m contains collinea points. Using containes and the fact that γ = 2, we conclude via the union bound that the pobability that thee exists an independent set of size m in H[R ] is at most C ) γ P p m 2 3 n m γep P m ) m ) m = 2 3 n e 2 3 n m. 1 2 Theefoe, w.h.p., R does not contain any independent sets of size m, povided 3 n m. As P 2n), we have m 2n) 20n 1 = 2n) 1 and thus 3/2 10 3/2 m 3 n n) /2 3 n = n n 3, fo a lage-enough constant. We want the last expession to tend to infinity, which holds if n 102. This, in tun, holds fo log n = 6 log log n. 5) Note that this is the only place in the poof whee we actually need to be of ode log 0.5 o1) n, and not log 1 o1) n, and the eason fo it is the facto 3 in 3). Now delete one point in R fom each collinea + 1)-tuple of R, obtaining the set R R. Clealy, R contains no collinea + 1)-tuples, and with high pobability, it has size 1 o 1)) p P by inequality 4)), and no independent sets of size m. Poject R on the plane so that no new collineaities appea. This implies that, with set by 5), thee exists a set R whose pojection in the plane satisfies these two conditions: H[R] does not contain independent sets of size at least m, and R has size 1 o1))p P. 9

11 Given the intege k, set the value of n so that k = 2 R 2p P n 1+o1). Then log 1 2 log log n 1 2 log log k 1 log, 2 which implies that log log n 2 3 log log k and that 3 = log n 6 log log n = log k 6 log log n log k 4 log log k. We have that R = k 2 k ) 1 log k 3. 6) 2 4 log log k 3.4 Lowe-bound on VC-dimension of the k-fold union induced by lines It emains to show a lowe-bound on the VC-dimension of the k-fold union of the set system induced by lines on R. In othe wods, given any S R, we want to show that thee exist k lines in the plane such that i) S lies in the union of these k lines, and ii) no point of R \ S lies on any of these lines. This would imply that R is shatteed by the k-fold union of lines in the plane, and thus the bound in inequality 6) gives a lowe-bound on the size of a shatteed set, which then gives the equied lowe-bound on the VC-dimension of k-fold union of the set system induced by lines in the plane. Take any subset S R. We constuct the equied set A of k lines iteatively. Initially set A =. If S m, then as shown above, thee is a collinea -tuple in S. This means that thee is a line l 1 that intesects S in points and does not intesect R \ S hee we need the fact that thee ae no + 1)-tuples in R! Add l 1 to A, emove the coveed points fom S and iteate as long as the emaining set still has size at least m. This can continue fo at most R = k 2 steps, afte which it must be that the size of the emaining set S is less than m. So fa, A k 2. When S < m R k, then we add one geneal position line pe point of S to A, which gives 3/2 an additional ok) lines. We have thus added at most k lines in total to A, and by constuction, these lines cove all the points of S and no point of R \ S. This holds fo any S R, and thus R is shatteed by the set system induced by the k-fold union of lines in the plane. This completes the poof. 10

12 4 Poof of Theoem 2. Poof. The poof will need the following lemma fom [12]: Lemma 6. Let n, d 2 be integes. Then thee exists a set B of d 2 n + 3)2n 2 axis-paallel boxes in R d such that fo any subset S B, one can find a 2 n 1 -element set Q of points in R d with the popety that i) Q B fo any B B \ S, and ii) Q B = fo any B S. By a standad lifting tansfom e.g., see [17, 12]), given a set B of boxes in R d, thee exists a function π : B R 2d mapping boxes in B to points in R 2d such that fo any point p R d, thee exists a coesponding half-space in R 2d, denoted by H p, such that a box B B contains p if and only if H p contains the point fb) R 2d. Fo any B B, set πb ) = {fb): B B }. Apply Lemma 6 with n = log k + 1 in R d/2 to get a set B of boxes in R d/2. Then P = πb) will be the equied point-set in R d ; i.e., we claim that P is shatteed by the set system induced by the k-fold union of half-spaces in R d. To see that, let P be any subset of P. Set S = π 1 P \ P ). By Lemma 6, thee exists a set Q of 2 n 1 = 2 log k k points in R d/2 such that each box in S contains no point of Q, and each box in B \ S contains at least one point of Q. We show that the set HP ) = {H p : p Q} of half-spaces sepaates P fom P. By the popety of the lifting map π ), each box in S contains no point of Q = each box in B \ S contains a point of Q = each point in P \ P is contained in no half-space of HP ), each point in P is contained in some half-space in HP ). In othe wods, the union of the half-spaces in HP ) contains pecisely the set P. As this is tue fo any P P, the k-fold union of half-spaces in R d shattes P. Finally, we have as desied. P = B = d 2 log k + 3)2 log k 2 = Ωd k log k), 11

13 5 Poof of Theoem 3. Poof. Set k = 1 2ɛ, and let d = VC-dimR). As VC-dimRk ) d 2ɛ f 1 2ɛ ), we can assume that X = d 2ɛ f 1 2ɛ ) and that it is shatteed by R 1 2ɛ. We will show that if N is an ɛ-net fo R, then N X 2 = d ) 1 4ɛ f. 2ɛ Suppose that N < X 2. Since X is shatteed by R 1 2ɛ, thee exists a set S R 1 2ɛ containing pecisely the elements in X \ N. In othe wods, we can find 1 2ɛ sets S 1,..., S 1 R such that 2ɛ ) S1 S 1 X = X\N. 2ɛ Each set S i contains no point of N and by the pigeonhole pinciple, one of them must have size at least X \ N 1 2ɛ X /2 1 2ɛ = ɛ X, and which is not hit by N. This contadicts the fact that N was an ɛ-net. 12

14 Refeences [1] Noga Alon. A non-linea lowe bound fo plana epsilon-nets. Discete & Computational Geomety, 472): , [2] Olivie Bachem, Maio Lucic, and Andeas Kause. Scalable and distibuted clusteing via lightweight coesets. CoRR, abs/ , [3] J. Balogh and J. Solymosi. On the numbe of points in geneal position in the plane. AXiv e-pints, Apil [4] A. Blume, A. Ehenfeucht, D. Haussle, and M. K. Wamuth. Leanability and the Vapnik- Chevonenkis dimension. J. ACM, 364): , Octobe [5] B. Chazelle. The Discepancy Method: Randomness and Complexity. Cambidge Univesity Pess, Cambidge, New Yok, [6] D. P. Dobkin and D. Gunopulos. Concept leaning with geometic hypotheses. In Poceedings of the Eighth Annual Confeence on Computational Leaning Theoy, COLT 95, pages , New Yok, NY, USA, ACM. [7] D. Eisenstat. k-fold unions of low-dimensional concept classes. Infomation Pocessing Lettes, ): , [8] D. Eisenstat and D. Angluin. The VC dimension of k-fold union. Infomation Pocessing Lettes, 1015): , [9] Dan Feldman and Michael Langbeg. A unified famewok fo appoximating and clusteing data. In Poceedings of the 43d ACM Symposium on Theoy of Computing, STOC 2011, San Jose, CA, USA, 6-8 June 2011, pages , [10] D. Haussle and E. Welzl. ε-nets and simplex ange queies. Discete & Computational Geomety, 2: , [11] H. Johnson. Definable families of finite Vapnik Chenonenkis dimension. PhD thesis, Univesity of Mayland, [12] A. Kupavskii, N. H. Mustafa, and J. Pach. New lowe bounds fo epsilon-nets. In 32nd Intenational Symposium on Computational Geomety SoCG), pages 54:1 54:16, [13] Jií Matousek. Lectues in Discete Geomety. Spinge-Velag, New Yok, NY, [14] N. H. Mustafa and K. Vaadaajan. Epsilon-appoximations and epsilon-nets. In J. E. Goodman, J. O Rouke, and C. D. Tóth, editos, Handbook of Discete and Computational Geomety. CRC Pess LLC, 2017, to appea. [15] J.E. Nymann. On the pobability that k positive integes ae elatively pime. Jounal of Numbe Theoy, 4: , [16] J. Pach and P. K. Agawal. Combinatoial Geomety. John Wiley & Sons, New Yok, NY, [17] J. Pach and G. Tados. Tight lowe bounds fo the size of epsilon-nets. In Symposium on Computational Geomety, pages ,

15 [18] D. Saxton and A. Thomason. Hypegaph containes. Inventiones mathematicae, pages 1 68,

On the ratio of maximum and minimum degree in maximal intersecting families

On the ratio of maximum and minimum degree in maximal intersecting families On the atio of maximum and minimum degee in maximal intesecting families Zoltán Lóánt Nagy Lale Özkahya Balázs Patkós Máté Vize Mach 6, 013 Abstact To study how balanced o unbalanced a maximal intesecting

More information

New problems in universal algebraic geometry illustrated by boolean equations

New problems in universal algebraic geometry illustrated by boolean equations New poblems in univesal algebaic geomety illustated by boolean equations axiv:1611.00152v2 [math.ra] 25 Nov 2016 Atem N. Shevlyakov Novembe 28, 2016 Abstact We discuss new poblems in univesal algebaic

More information

On the ratio of maximum and minimum degree in maximal intersecting families

On the ratio of maximum and minimum degree in maximal intersecting families On the atio of maximum and minimum degee in maximal intesecting families Zoltán Lóánt Nagy Lale Özkahya Balázs Patkós Máté Vize Septembe 5, 011 Abstact To study how balanced o unbalanced a maximal intesecting

More information

COLLAPSING WALLS THEOREM

COLLAPSING WALLS THEOREM COLLAPSING WALLS THEOREM IGOR PAK AND ROM PINCHASI Abstact. Let P R 3 be a pyamid with the base a convex polygon Q. We show that when othe faces ae collapsed (otated aound the edges onto the plane spanned

More information

Quasi-Randomness and the Distribution of Copies of a Fixed Graph

Quasi-Randomness and the Distribution of Copies of a Fixed Graph Quasi-Randomness and the Distibution of Copies of a Fixed Gaph Asaf Shapia Abstact We show that if a gaph G has the popety that all subsets of vetices of size n/4 contain the coect numbe of tiangles one

More information

A solution to a problem of Grünbaum and Motzkin and of Erdős and Purdy about bichromatic configurations of points in the plane

A solution to a problem of Grünbaum and Motzkin and of Erdős and Purdy about bichromatic configurations of points in the plane A solution to a poblem of Günbaum and Motzkin and of Edős and Pudy about bichomatic configuations of points in the plane Rom Pinchasi July 29, 2012 Abstact Let P be a set of n blue points in the plane,

More information

ON INDEPENDENT SETS IN PURELY ATOMIC PROBABILITY SPACES WITH GEOMETRIC DISTRIBUTION. 1. Introduction. 1 r r. r k for every set E A, E \ {0},

ON INDEPENDENT SETS IN PURELY ATOMIC PROBABILITY SPACES WITH GEOMETRIC DISTRIBUTION. 1. Introduction. 1 r r. r k for every set E A, E \ {0}, ON INDEPENDENT SETS IN PURELY ATOMIC PROBABILITY SPACES WITH GEOMETRIC DISTRIBUTION E. J. IONASCU and A. A. STANCU Abstact. We ae inteested in constucting concete independent events in puely atomic pobability

More information

NOTE. Some New Bounds for Cover-Free Families

NOTE. Some New Bounds for Cover-Free Families Jounal of Combinatoial Theoy, Seies A 90, 224234 (2000) doi:10.1006jcta.1999.3036, available online at http:.idealibay.com on NOTE Some Ne Bounds fo Cove-Fee Families D. R. Stinson 1 and R. Wei Depatment

More information

ON THE INVERSE SIGNED TOTAL DOMINATION NUMBER IN GRAPHS. D.A. Mojdeh and B. Samadi

ON THE INVERSE SIGNED TOTAL DOMINATION NUMBER IN GRAPHS. D.A. Mojdeh and B. Samadi Opuscula Math. 37, no. 3 (017), 447 456 http://dx.doi.og/10.7494/opmath.017.37.3.447 Opuscula Mathematica ON THE INVERSE SIGNED TOTAL DOMINATION NUMBER IN GRAPHS D.A. Mojdeh and B. Samadi Communicated

More information

Research Article On Alzer and Qiu s Conjecture for Complete Elliptic Integral and Inverse Hyperbolic Tangent Function

Research Article On Alzer and Qiu s Conjecture for Complete Elliptic Integral and Inverse Hyperbolic Tangent Function Abstact and Applied Analysis Volume 011, Aticle ID 697547, 7 pages doi:10.1155/011/697547 Reseach Aticle On Alze and Qiu s Conjectue fo Complete Elliptic Integal and Invese Hypebolic Tangent Function Yu-Ming

More information

Multiple Criteria Secretary Problem: A New Approach

Multiple Criteria Secretary Problem: A New Approach J. Stat. Appl. Po. 3, o., 9-38 (04 9 Jounal of Statistics Applications & Pobability An Intenational Jounal http://dx.doi.og/0.785/jsap/0303 Multiple Citeia Secetay Poblem: A ew Appoach Alaka Padhye, and

More information

ONE-POINT CODES USING PLACES OF HIGHER DEGREE

ONE-POINT CODES USING PLACES OF HIGHER DEGREE ONE-POINT CODES USING PLACES OF HIGHER DEGREE GRETCHEN L. MATTHEWS AND TODD W. MICHEL DEPARTMENT OF MATHEMATICAL SCIENCES CLEMSON UNIVERSITY CLEMSON, SC 29634-0975 U.S.A. E-MAIL: GMATTHE@CLEMSON.EDU, TMICHEL@CLEMSON.EDU

More information

The Congestion of n-cube Layout on a Rectangular Grid S.L. Bezrukov J.D. Chavez y L.H. Harper z M. Rottger U.-P. Schroeder Abstract We consider the pr

The Congestion of n-cube Layout on a Rectangular Grid S.L. Bezrukov J.D. Chavez y L.H. Harper z M. Rottger U.-P. Schroeder Abstract We consider the pr The Congestion of n-cube Layout on a Rectangula Gid S.L. Bezukov J.D. Chavez y L.H. Hape z M. Rottge U.-P. Schoede Abstact We conside the poblem of embedding the n-dimensional cube into a ectangula gid

More information

Lecture 18: Graph Isomorphisms

Lecture 18: Graph Isomorphisms INFR11102: Computational Complexity 22/11/2018 Lectue: Heng Guo Lectue 18: Gaph Isomophisms 1 An Athu-Melin potocol fo GNI Last time we gave a simple inteactive potocol fo GNI with pivate coins. We will

More information

arxiv: v1 [math.co] 4 May 2017

arxiv: v1 [math.co] 4 May 2017 On The Numbe Of Unlabeled Bipatite Gaphs Abdullah Atmaca and A Yavuz Ouç axiv:7050800v [mathco] 4 May 207 Abstact This pape solves a poblem that was stated by M A Haison in 973 [] This poblem, that has

More information

On decompositions of complete multipartite graphs into the union of two even cycles

On decompositions of complete multipartite graphs into the union of two even cycles On decompositions of complete multipatite gaphs into the union of two even cycles A. Su, J. Buchanan, R. C. Bunge, S. I. El-Zanati, E. Pelttai, G. Rasmuson, E. Spaks, S. Tagais Depatment of Mathematics

More information

Chapter 3: Theory of Modular Arithmetic 38

Chapter 3: Theory of Modular Arithmetic 38 Chapte 3: Theoy of Modula Aithmetic 38 Section D Chinese Remainde Theoem By the end of this section you will be able to pove the Chinese Remainde Theoem apply this theoem to solve simultaneous linea conguences

More information

Duality between Statical and Kinematical Engineering Systems

Duality between Statical and Kinematical Engineering Systems Pape 00, Civil-Comp Ltd., Stiling, Scotland Poceedings of the Sixth Intenational Confeence on Computational Stuctues Technology, B.H.V. Topping and Z. Bittna (Editos), Civil-Comp Pess, Stiling, Scotland.

More information

Unobserved Correlation in Ascending Auctions: Example And Extensions

Unobserved Correlation in Ascending Auctions: Example And Extensions Unobseved Coelation in Ascending Auctions: Example And Extensions Daniel Quint Univesity of Wisconsin Novembe 2009 Intoduction In pivate-value ascending auctions, the winning bidde s willingness to pay

More information

Central Coverage Bayes Prediction Intervals for the Generalized Pareto Distribution

Central Coverage Bayes Prediction Intervals for the Generalized Pareto Distribution Statistics Reseach Lettes Vol. Iss., Novembe Cental Coveage Bayes Pediction Intevals fo the Genealized Paeto Distibution Gyan Pakash Depatment of Community Medicine S. N. Medical College, Aga, U. P., India

More information

Secret Exponent Attacks on RSA-type Schemes with Moduli N = p r q

Secret Exponent Attacks on RSA-type Schemes with Moduli N = p r q Secet Exponent Attacks on RSA-type Schemes with Moduli N = p q Alexande May Faculty of Compute Science, Electical Engineeing and Mathematics Univesity of Padebon 33102 Padebon, Gemany alexx@uni-padebon.de

More information

q i i=1 p i ln p i Another measure, which proves a useful benchmark in our analysis, is the chi squared divergence of p, q, which is defined by

q i i=1 p i ln p i Another measure, which proves a useful benchmark in our analysis, is the chi squared divergence of p, q, which is defined by CSISZÁR f DIVERGENCE, OSTROWSKI S INEQUALITY AND MUTUAL INFORMATION S. S. DRAGOMIR, V. GLUŠČEVIĆ, AND C. E. M. PEARCE Abstact. The Ostowski integal inequality fo an absolutely continuous function is used

More information

Math 301: The Erdős-Stone-Simonovitz Theorem and Extremal Numbers for Bipartite Graphs

Math 301: The Erdős-Stone-Simonovitz Theorem and Extremal Numbers for Bipartite Graphs Math 30: The Edős-Stone-Simonovitz Theoem and Extemal Numbes fo Bipatite Gaphs May Radcliffe The Edős-Stone-Simonovitz Theoem Recall, in class we poved Tuán s Gaph Theoem, namely Theoem Tuán s Theoem Let

More information

Approximating Maximum Diameter-Bounded Subgraph in Unit Disk Graphs

Approximating Maximum Diameter-Bounded Subgraph in Unit Disk Graphs Appoximating Maximum Diamete-Bounded Subgaph in Unit Disk Gaphs A. Kaim Abu-Affash 1 Softwae Engineeing Depatment, Shamoon College of Engineeing Bee-Sheva 84100, Isael abuaa1@sce.ac.il Paz Cami Depatment

More information

10/04/18. P [P(x)] 1 negl(n).

10/04/18. P [P(x)] 1 negl(n). Mastemath, Sping 208 Into to Lattice lgs & Cypto Lectue 0 0/04/8 Lectues: D. Dadush, L. Ducas Scibe: K. de Boe Intoduction In this lectue, we will teat two main pats. Duing the fist pat we continue the

More information

Lifting Private Information Retrieval from Two to any Number of Messages

Lifting Private Information Retrieval from Two to any Number of Messages Lifting Pivate Infomation Retieval fom Two to any umbe of Messages Rafael G.L. D Oliveia, Salim El Rouayheb ECE, Rutges Univesity, Piscataway, J Emails: d746@scaletmail.utges.edu, salim.elouayheb@utges.edu

More information

Stanford University CS259Q: Quantum Computing Handout 8 Luca Trevisan October 18, 2012

Stanford University CS259Q: Quantum Computing Handout 8 Luca Trevisan October 18, 2012 Stanfod Univesity CS59Q: Quantum Computing Handout 8 Luca Tevisan Octobe 8, 0 Lectue 8 In which we use the quantum Fouie tansfom to solve the peiod-finding poblem. The Peiod Finding Poblem Let f : {0,...,

More information

Journal of Inequalities in Pure and Applied Mathematics

Journal of Inequalities in Pure and Applied Mathematics Jounal of Inequalities in Pue and Applied Mathematics COEFFICIENT INEQUALITY FOR A FUNCTION WHOSE DERIVATIVE HAS A POSITIVE REAL PART S. ABRAMOVICH, M. KLARIČIĆ BAKULA AND S. BANIĆ Depatment of Mathematics

More information

Brief summary of functional analysis APPM 5440 Fall 2014 Applied Analysis

Brief summary of functional analysis APPM 5440 Fall 2014 Applied Analysis Bief summay of functional analysis APPM 5440 Fall 014 Applied Analysis Stephen Becke, stephen.becke@coloado.edu Standad theoems. When necessay, I used Royden s and Keyzsig s books as a efeence. Vesion

More information

Surveillance Points in High Dimensional Spaces

Surveillance Points in High Dimensional Spaces Société de Calcul Mathématique SA Tools fo decision help since 995 Suveillance Points in High Dimensional Spaces by Benad Beauzamy Januay 06 Abstact Let us conside any compute softwae, elying upon a lage

More information

Lecture 8 - Gauss s Law

Lecture 8 - Gauss s Law Lectue 8 - Gauss s Law A Puzzle... Example Calculate the potential enegy, pe ion, fo an infinite 1D ionic cystal with sepaation a; that is, a ow of equally spaced chages of magnitude e and altenating sign.

More information

An intersection theorem for four sets

An intersection theorem for four sets An intesection theoem fo fou sets Dhuv Mubayi Novembe 22, 2006 Abstact Fix integes n, 4 and let F denote a family of -sets of an n-element set Suppose that fo evey fou distinct A, B, C, D F with A B C

More information

1 Explicit Explore or Exploit (E 3 ) Algorithm

1 Explicit Explore or Exploit (E 3 ) Algorithm 2.997 Decision-Making in Lage-Scale Systems Mach 3 MIT, Sping 2004 Handout #2 Lectue Note 9 Explicit Exploe o Exploit (E 3 ) Algoithm Last lectue, we studied the Q-leaning algoithm: [ ] Q t+ (x t, a t

More information

SOME GENERAL NUMERICAL RADIUS INEQUALITIES FOR THE OFF-DIAGONAL PARTS OF 2 2 OPERATOR MATRICES

SOME GENERAL NUMERICAL RADIUS INEQUALITIES FOR THE OFF-DIAGONAL PARTS OF 2 2 OPERATOR MATRICES italian jounal of pue and applied mathematics n. 35 015 (433 44) 433 SOME GENERAL NUMERICAL RADIUS INEQUALITIES FOR THE OFF-DIAGONAL PARTS OF OPERATOR MATRICES Watheq Bani-Domi Depatment of Mathematics

More information

On the Poisson Approximation to the Negative Hypergeometric Distribution

On the Poisson Approximation to the Negative Hypergeometric Distribution BULLETIN of the Malaysian Mathematical Sciences Society http://mathusmmy/bulletin Bull Malays Math Sci Soc (2) 34(2) (2011), 331 336 On the Poisson Appoximation to the Negative Hypegeometic Distibution

More information

6 PROBABILITY GENERATING FUNCTIONS

6 PROBABILITY GENERATING FUNCTIONS 6 PROBABILITY GENERATING FUNCTIONS Cetain deivations pesented in this couse have been somewhat heavy on algeba. Fo example, detemining the expectation of the Binomial distibution (page 5.1 tuned out to

More information

MULTILAYER PERCEPTRONS

MULTILAYER PERCEPTRONS Last updated: Nov 26, 2012 MULTILAYER PERCEPTRONS Outline 2 Combining Linea Classifies Leaning Paametes Outline 3 Combining Linea Classifies Leaning Paametes Implementing Logical Relations 4 AND and OR

More information

Lecture 28: Convergence of Random Variables and Related Theorems

Lecture 28: Convergence of Random Variables and Related Theorems EE50: Pobability Foundations fo Electical Enginees July-Novembe 205 Lectue 28: Convegence of Random Vaiables and Related Theoems Lectue:. Kishna Jagannathan Scibe: Gopal, Sudhasan, Ajay, Swamy, Kolla An

More information

Deterministic vs Non-deterministic Graph Property Testing

Deterministic vs Non-deterministic Graph Property Testing Deteministic vs Non-deteministic Gaph Popety Testing Lio Gishboline Asaf Shapia Abstact A gaph popety P is said to be testable if one can check whethe a gaph is close o fa fom satisfying P using few andom

More information

SUFFICIENT CONDITIONS FOR MAXIMALLY EDGE-CONNECTED AND SUPER-EDGE-CONNECTED GRAPHS DEPENDING ON THE CLIQUE NUMBER

SUFFICIENT CONDITIONS FOR MAXIMALLY EDGE-CONNECTED AND SUPER-EDGE-CONNECTED GRAPHS DEPENDING ON THE CLIQUE NUMBER Discussiones Mathematicae Gaph Theoy 39 (019) 567 573 doi:10.7151/dmgt.096 SUFFICIENT CONDITIONS FOR MAXIMALLY EDGE-CONNECTED AND SUPER-EDGE-CONNECTED GRAPHS DEPENDING ON THE CLIQUE NUMBER Lutz Volkmann

More information

Geometry of the homogeneous and isotropic spaces

Geometry of the homogeneous and isotropic spaces Geomety of the homogeneous and isotopic spaces H. Sonoda Septembe 2000; last evised Octobe 2009 Abstact We summaize the aspects of the geomety of the homogeneous and isotopic spaces which ae most elevant

More information

Physics 2A Chapter 10 - Moment of Inertia Fall 2018

Physics 2A Chapter 10 - Moment of Inertia Fall 2018 Physics Chapte 0 - oment of netia Fall 08 The moment of inetia of a otating object is a measue of its otational inetia in the same way that the mass of an object is a measue of its inetia fo linea motion.

More information

4/18/2005. Statistical Learning Theory

4/18/2005. Statistical Learning Theory Statistical Leaning Theoy Statistical Leaning Theoy A model of supevised leaning consists of: a Envionment - Supplying a vecto x with a fixed but unknown pdf F x (x b Teache. It povides a desied esponse

More information

The height of minimal Hilbert bases

The height of minimal Hilbert bases 1 The height of minimal Hilbet bases Matin Henk and Robet Weismantel Abstact Fo an integal polyhedal cone C = pos{a 1,..., a m, a i Z d, a subset BC) C Z d is called a minimal Hilbet basis of C iff i)

More information

Fall 2014 Randomized Algorithms Oct 8, Lecture 3

Fall 2014 Randomized Algorithms Oct 8, Lecture 3 Fall 204 Randomized Algoithms Oct 8, 204 Lectue 3 Pof. Fiedich Eisenband Scibes: Floian Tamè In this lectue we will be concened with linea pogamming, in paticula Clakson s Las Vegas algoithm []. The main

More information

arxiv: v1 [math.co] 1 Apr 2011

arxiv: v1 [math.co] 1 Apr 2011 Weight enumeation of codes fom finite spaces Relinde Juius Octobe 23, 2018 axiv:1104.0172v1 [math.co] 1 Ap 2011 Abstact We study the genealized and extended weight enumeato of the - ay Simplex code and

More information

Probablistically Checkable Proofs

Probablistically Checkable Proofs Lectue 12 Pobablistically Checkable Poofs May 13, 2004 Lectue: Paul Beame Notes: Chis Re 12.1 Pobablisitically Checkable Poofs Oveview We know that IP = PSPACE. This means thee is an inteactive potocol

More information

9.1 The multiplicative group of a finite field. Theorem 9.1. The multiplicative group F of a finite field is cyclic.

9.1 The multiplicative group of a finite field. Theorem 9.1. The multiplicative group F of a finite field is cyclic. Chapte 9 Pimitive Roots 9.1 The multiplicative goup of a finite fld Theoem 9.1. The multiplicative goup F of a finite fld is cyclic. Remak: In paticula, if p is a pime then (Z/p) is cyclic. In fact, this

More information

Measure Estimates of Nodal Sets of Polyharmonic Functions

Measure Estimates of Nodal Sets of Polyharmonic Functions Chin. Ann. Math. Se. B 39(5), 08, 97 93 DOI: 0.007/s40-08-004-6 Chinese Annals of Mathematics, Seies B c The Editoial Office of CAM and Spinge-Velag Belin Heidelbeg 08 Measue Estimates of Nodal Sets of

More information

titrrvers:rtt t>1 NO~~H CAROLINA

titrrvers:rtt t>1 NO~~H CAROLINA titvers:tt t>1 NO~~H CAROLINA Depatment of statistics Chapel Hill, N. C. ON A BOUN.D USEFUL IN THE THEORY OF FACTORIAL DESIGNS AND ERROR CORRECTING CODES by R. C. Bose and J. N. Sivastava Apil 1963 Gant

More information

On a quantity that is analogous to potential and a theorem that relates to it

On a quantity that is analogous to potential and a theorem that relates to it Su une quantité analogue au potential et su un théoème y elatif C R Acad Sci 7 (87) 34-39 On a quantity that is analogous to potential and a theoem that elates to it By R CLAUSIUS Tanslated by D H Delphenich

More information

A Bijective Approach to the Permutational Power of a Priority Queue

A Bijective Approach to the Permutational Power of a Priority Queue A Bijective Appoach to the Pemutational Powe of a Pioity Queue Ia M. Gessel Kuang-Yeh Wang Depatment of Mathematics Bandeis Univesity Waltham, MA 02254-9110 Abstact A pioity queue tansfoms an input pemutation

More information

Vanishing lines in generalized Adams spectral sequences are generic

Vanishing lines in generalized Adams spectral sequences are generic ISSN 364-0380 (on line) 465-3060 (pinted) 55 Geomety & Topology Volume 3 (999) 55 65 Published: 2 July 999 G G G G T T T G T T T G T G T GG TT G G G G GG T T T TT Vanishing lines in genealized Adams spectal

More information

Using Laplace Transform to Evaluate Improper Integrals Chii-Huei Yu

Using Laplace Transform to Evaluate Improper Integrals Chii-Huei Yu Available at https://edupediapublicationsog/jounals Volume 3 Issue 4 Febuay 216 Using Laplace Tansfom to Evaluate Impope Integals Chii-Huei Yu Depatment of Infomation Technology, Nan Jeon Univesity of

More information

A Multivariate Normal Law for Turing s Formulae

A Multivariate Normal Law for Turing s Formulae A Multivaiate Nomal Law fo Tuing s Fomulae Zhiyi Zhang Depatment of Mathematics and Statistics Univesity of Noth Caolina at Chalotte Chalotte, NC 28223 Abstact This pape establishes a sufficient condition

More information

To Feel a Force Chapter 7 Static equilibrium - torque and friction

To Feel a Force Chapter 7 Static equilibrium - torque and friction To eel a oce Chapte 7 Chapte 7: Static fiction, toque and static equilibium A. Review of foce vectos Between the eath and a small mass, gavitational foces of equal magnitude and opposite diection act on

More information

Functions Defined on Fuzzy Real Numbers According to Zadeh s Extension

Functions Defined on Fuzzy Real Numbers According to Zadeh s Extension Intenational Mathematical Foum, 3, 2008, no. 16, 763-776 Functions Defined on Fuzzy Real Numbes Accoding to Zadeh s Extension Oma A. AbuAaqob, Nabil T. Shawagfeh and Oma A. AbuGhneim 1 Mathematics Depatment,

More information

EM Boundary Value Problems

EM Boundary Value Problems EM Bounday Value Poblems 10/ 9 11/ By Ilekta chistidi & Lee, Seung-Hyun A. Geneal Desciption : Maxwell Equations & Loentz Foce We want to find the equations of motion of chaged paticles. The way to do

More information

Upper Bounds for Tura n Numbers. Alexander Sidorenko

Upper Bounds for Tura n Numbers. Alexander Sidorenko jounal of combinatoial theoy, Seies A 77, 134147 (1997) aticle no. TA962739 Uppe Bounds fo Tua n Numbes Alexande Sidoenko Couant Institute of Mathematical Sciences, New Yok Univesity, 251 Mece Steet, New

More information

Method for Approximating Irrational Numbers

Method for Approximating Irrational Numbers Method fo Appoximating Iational Numbes Eic Reichwein Depatment of Physics Univesity of Califonia, Santa Cuz June 6, 0 Abstact I will put foth an algoithm fo poducing inceasingly accuate ational appoximations

More information

Construction and Analysis of Boolean Functions of 2t + 1 Variables with Maximum Algebraic Immunity

Construction and Analysis of Boolean Functions of 2t + 1 Variables with Maximum Algebraic Immunity Constuction and Analysis of Boolean Functions of 2t + 1 Vaiables with Maximum Algebaic Immunity Na Li and Wen-Feng Qi Depatment of Applied Mathematics, Zhengzhou Infomation Engineeing Univesity, Zhengzhou,

More information

Fractional Zero Forcing via Three-color Forcing Games

Fractional Zero Forcing via Three-color Forcing Games Factional Zeo Focing via Thee-colo Focing Games Leslie Hogben Kevin F. Palmowski David E. Robeson Michael Young May 13, 2015 Abstact An -fold analogue of the positive semidefinite zeo focing pocess that

More information

Additive Approximation for Edge-Deletion Problems

Additive Approximation for Edge-Deletion Problems Additive Appoximation fo Edge-Deletion Poblems Noga Alon Asaf Shapia Benny Sudakov Abstact A gaph popety is monotone if it is closed unde emoval of vetices and edges. In this pape we conside the following

More information

The Erdős-Hajnal conjecture for rainbow triangles

The Erdős-Hajnal conjecture for rainbow triangles The Edős-Hajnal conjectue fo ainbow tiangles Jacob Fox Andey Ginshpun János Pach Abstact We pove that evey 3-coloing of the edges of the complete gaph on n vetices without a ainbow tiangle contains a set

More information

Absorption Rate into a Small Sphere for a Diffusing Particle Confined in a Large Sphere

Absorption Rate into a Small Sphere for a Diffusing Particle Confined in a Large Sphere Applied Mathematics, 06, 7, 709-70 Published Online Apil 06 in SciRes. http://www.scip.og/jounal/am http://dx.doi.og/0.46/am.06.77065 Absoption Rate into a Small Sphee fo a Diffusing Paticle Confined in

More information

INFORMATION GEOMETRY OF PROPAGATION ALGORITHMS AND APPROXIMATE INFERENCE

INFORMATION GEOMETRY OF PROPAGATION ALGORITHMS AND APPROXIMATE INFERENCE 2nd Intenational Symposium on Infomation Geomety and its Applications Decembe 12-16, 2005, Tokyo Pages 000 000 INFORMATION GEOMETRY OF PROPAGATION ALGORITHMS AND APPROXIMATE INFERENCE SHIRO IKEDA 1 intoduction

More information

arxiv: v1 [physics.pop-ph] 3 Jun 2013

arxiv: v1 [physics.pop-ph] 3 Jun 2013 A note on the electostatic enegy of two point chages axiv:1306.0401v1 [physics.pop-ph] 3 Jun 013 A C Tot Instituto de Física Univesidade Fedeal do io de Janeio Caixa Postal 68.58; CEP 1941-97 io de Janeio,

More information

A thermodynamic degree of freedom solution to the galaxy cluster problem of MOND. Abstract

A thermodynamic degree of freedom solution to the galaxy cluster problem of MOND. Abstract A themodynamic degee of feedom solution to the galaxy cluste poblem of MOND E.P.J. de Haas (Paul) Nijmegen, The Nethelands (Dated: Octobe 23, 2015) Abstact In this pape I discus the degee of feedom paamete

More information

Enumerating permutation polynomials

Enumerating permutation polynomials Enumeating pemutation polynomials Theodoulos Gaefalakis a,1, Giogos Kapetanakis a,, a Depatment of Mathematics and Applied Mathematics, Univesity of Cete, 70013 Heaklion, Geece Abstact We conside thoblem

More information

AQI: Advanced Quantum Information Lecture 2 (Module 4): Order finding and factoring algorithms February 20, 2013

AQI: Advanced Quantum Information Lecture 2 (Module 4): Order finding and factoring algorithms February 20, 2013 AQI: Advanced Quantum Infomation Lectue 2 (Module 4): Ode finding and factoing algoithms Febuay 20, 203 Lectue: D. Mak Tame (email: m.tame@impeial.ac.uk) Intoduction In the last lectue we looked at the

More information

ON SPARSELY SCHEMMEL TOTIENT NUMBERS. Colin Defant 1 Department of Mathematics, University of Florida, Gainesville, Florida

ON SPARSELY SCHEMMEL TOTIENT NUMBERS. Colin Defant 1 Department of Mathematics, University of Florida, Gainesville, Florida #A8 INTEGERS 5 (205) ON SPARSEL SCHEMMEL TOTIENT NUMBERS Colin Defant Depatment of Mathematics, Univesity of Floida, Gainesville, Floida cdefant@ufl.edu Received: 7/30/4, Revised: 2/23/4, Accepted: 4/26/5,

More information

Semicanonical basis generators of the cluster algebra of type A (1)

Semicanonical basis generators of the cluster algebra of type A (1) Semicanonical basis geneatos of the cluste algeba of type A (1 1 Andei Zelevinsky Depatment of Mathematics Notheasten Univesity, Boston, USA andei@neu.edu Submitted: Jul 7, 006; Accepted: Dec 3, 006; Published:

More information

Hypothesis Test and Confidence Interval for the Negative Binomial Distribution via Coincidence: A Case for Rare Events

Hypothesis Test and Confidence Interval for the Negative Binomial Distribution via Coincidence: A Case for Rare Events Intenational Jounal of Contempoay Mathematical Sciences Vol. 12, 2017, no. 5, 243-253 HIKARI Ltd, www.m-hikai.com https://doi.og/10.12988/ijcms.2017.7728 Hypothesis Test and Confidence Inteval fo the Negative

More information

3.1 Random variables

3.1 Random variables 3 Chapte III Random Vaiables 3 Random vaiables A sample space S may be difficult to descibe if the elements of S ae not numbes discuss how we can use a ule by which an element s of S may be associated

More information

A NEW VARIABLE STIFFNESS SPRING USING A PRESTRESSED MECHANISM

A NEW VARIABLE STIFFNESS SPRING USING A PRESTRESSED MECHANISM Poceedings of the ASME 2010 Intenational Design Engineeing Technical Confeences & Computes and Infomation in Engineeing Confeence IDETC/CIE 2010 August 15-18, 2010, Monteal, Quebec, Canada DETC2010-28496

More information

PROBLEM SET #1 SOLUTIONS by Robert A. DiStasio Jr.

PROBLEM SET #1 SOLUTIONS by Robert A. DiStasio Jr. POBLM S # SOLUIONS by obet A. DiStasio J. Q. he Bon-Oppenheime appoximation is the standad way of appoximating the gound state of a molecula system. Wite down the conditions that detemine the tonic and

More information

THE MAXIMUM SIZE OF A PARTIAL SPREAD II: UPPER BOUNDS

THE MAXIMUM SIZE OF A PARTIAL SPREAD II: UPPER BOUNDS THE MAXIMUM SIZE OF A PARTIAL SPREAD II: UPPER BOUNDS ESMERALDA NĂSTASE MATHEMATICS DEPARTMENT XAVIER UNIVERSITY CINCINNATI, OHIO 4507, USA PAPA SISSOKHO MATHEMATICS DEPARTMENT ILLINOIS STATE UNIVERSITY

More information

MAGNETIC FIELD AROUND TWO SEPARATED MAGNETIZING COILS

MAGNETIC FIELD AROUND TWO SEPARATED MAGNETIZING COILS The 8 th Intenational Confeence of the Slovenian Society fo Non-Destuctive Testing»pplication of Contempoay Non-Destuctive Testing in Engineeing«Septembe 1-3, 5, Potoož, Slovenia, pp. 17-1 MGNETIC FIELD

More information

Identification of the degradation of railway ballast under a concrete sleeper

Identification of the degradation of railway ballast under a concrete sleeper Identification of the degadation of ailway ballast unde a concete sleepe Qin Hu 1) and Heung Fai Lam ) 1), ) Depatment of Civil and Achitectual Engineeing, City Univesity of Hong Kong, Hong Kong SAR, China.

More information

β β β β β B B B B o (i) (ii) (iii)

β β β β β B B B B o (i) (ii) (iii) Output-Sensitive Algoithms fo Unifom Patitions of Points Pankaj K. Agawal y Binay K. Bhattachaya z Sandeep Sen x Octobe 19, 1999 Abstact We conside the following one and two-dimensional bucketing poblems:

More information

Chapter 5 Force and Motion

Chapter 5 Force and Motion Chapte 5 Foce and Motion In Chaptes 2 and 4 we have studied kinematics, i.e., we descibed the motion of objects using paametes such as the position vecto, velocity, and acceleation without any insights

More information

working pages for Paul Richards class notes; do not copy or circulate without permission from PGR 2004/11/3 10:50

working pages for Paul Richards class notes; do not copy or circulate without permission from PGR 2004/11/3 10:50 woking pages fo Paul Richads class notes; do not copy o ciculate without pemission fom PGR 2004/11/3 10:50 CHAPTER7 Solid angle, 3D integals, Gauss s Theoem, and a Delta Function We define the solid angle,

More information

Solution to HW 3, Ma 1a Fall 2016

Solution to HW 3, Ma 1a Fall 2016 Solution to HW 3, Ma a Fall 206 Section 2. Execise 2: Let C be a subset of the eal numbes consisting of those eal numbes x having the popety that evey digit in the decimal expansion of x is, 3, 5, o 7.

More information

Relating Branching Program Size and. Formula Size over the Full Binary Basis. FB Informatik, LS II, Univ. Dortmund, Dortmund, Germany

Relating Branching Program Size and. Formula Size over the Full Binary Basis. FB Informatik, LS II, Univ. Dortmund, Dortmund, Germany Relating Banching Pogam Size and omula Size ove the ull Binay Basis Matin Saueho y Ingo Wegene y Ralph Wechne z y B Infomatik, LS II, Univ. Dotmund, 44 Dotmund, Gemany z ankfut, Gemany sauehof/wegene@ls.cs.uni-dotmund.de

More information

6 Matrix Concentration Bounds

6 Matrix Concentration Bounds 6 Matix Concentation Bounds Concentation bounds ae inequalities that bound pobabilities of deviations by a andom vaiable fom some value, often its mean. Infomally, they show the pobability that a andom

More information

KOEBE DOMAINS FOR THE CLASSES OF FUNCTIONS WITH RANGES INCLUDED IN GIVEN SETS

KOEBE DOMAINS FOR THE CLASSES OF FUNCTIONS WITH RANGES INCLUDED IN GIVEN SETS Jounal of Applied Analysis Vol. 14, No. 1 2008), pp. 43 52 KOEBE DOMAINS FOR THE CLASSES OF FUNCTIONS WITH RANGES INCLUDED IN GIVEN SETS L. KOCZAN and P. ZAPRAWA Received Mach 12, 2007 and, in evised fom,

More information

Conservative Averaging Method and its Application for One Heat Conduction Problem

Conservative Averaging Method and its Application for One Heat Conduction Problem Poceedings of the 4th WSEAS Int. Conf. on HEAT TRANSFER THERMAL ENGINEERING and ENVIRONMENT Elounda Geece August - 6 (pp6-) Consevative Aveaging Method and its Application fo One Heat Conduction Poblem

More information

Compactly Supported Radial Basis Functions

Compactly Supported Radial Basis Functions Chapte 4 Compactly Suppoted Radial Basis Functions As we saw ealie, compactly suppoted functions Φ that ae tuly stictly conditionally positive definite of ode m > do not exist The compact suppot automatically

More information

STUDY OF SOLUTIONS OF LOGARITHMIC ORDER TO HIGHER ORDER LINEAR DIFFERENTIAL-DIFFERENCE EQUATIONS WITH COEFFICIENTS HAVING THE SAME LOGARITHMIC ORDER

STUDY OF SOLUTIONS OF LOGARITHMIC ORDER TO HIGHER ORDER LINEAR DIFFERENTIAL-DIFFERENCE EQUATIONS WITH COEFFICIENTS HAVING THE SAME LOGARITHMIC ORDER UNIVERSITATIS IAGELLONICAE ACTA MATHEMATICA doi: 104467/20843828AM170027078 542017, 15 32 STUDY OF SOLUTIONS OF LOGARITHMIC ORDER TO HIGHER ORDER LINEAR DIFFERENTIAL-DIFFERENCE EQUATIONS WITH COEFFICIENTS

More information

Pearson s Chi-Square Test Modifications for Comparison of Unweighted and Weighted Histograms and Two Weighted Histograms

Pearson s Chi-Square Test Modifications for Comparison of Unweighted and Weighted Histograms and Two Weighted Histograms Peason s Chi-Squae Test Modifications fo Compaison of Unweighted and Weighted Histogams and Two Weighted Histogams Univesity of Akueyi, Bogi, v/noduslód, IS-6 Akueyi, Iceland E-mail: nikolai@unak.is Two

More information

Temporal-Difference Learning

Temporal-Difference Learning .997 Decision-Making in Lage-Scale Systems Mach 17 MIT, Sping 004 Handout #17 Lectue Note 13 1 Tempoal-Diffeence Leaning We now conside the poblem of computing an appopiate paamete, so that, given an appoximation

More information

New lower bounds for the independence number of sparse graphs and hypergraphs

New lower bounds for the independence number of sparse graphs and hypergraphs New lowe bounds fo the independence numbe of spase gaphs and hypegaphs Kunal Dutta, Dhuv Mubayi, and C.R. Subamanian May 23, 202 Abstact We obtain new lowe bounds fo the independence numbe of K -fee gaphs

More information

arxiv: v1 [math.nt] 12 May 2017

arxiv: v1 [math.nt] 12 May 2017 SEQUENCES OF CONSECUTIVE HAPPY NUMBERS IN NEGATIVE BASES HELEN G. GRUNDMAN AND PAMELA E. HARRIS axiv:1705.04648v1 [math.nt] 12 May 2017 ABSTRACT. Fo b 2 and e 2, let S e,b : Z Z 0 be the function taking

More information

Analysis of simple branching trees with TI-92

Analysis of simple branching trees with TI-92 Analysis of simple banching tees with TI-9 Dušan Pagon, Univesity of Maibo, Slovenia Abstact. In the complex plane we stat at the cente of the coodinate system with a vetical segment of the length one

More information

BLACK HOLES IN STRING THEORY

BLACK HOLES IN STRING THEORY Black holes in sting theoy N Sadikaj & A Duka Pape pesented in 1 -st Intenational Scientific Confeence on Pofessional Sciences, Alexande Moisiu Univesity, Dues Novembe 016 BLACK OLES IN STRING TEORY NDRIÇIM

More information

MASSACHUSETTS INSTITUTE OF TECHNOLOGY Physics Department. Problem Set 10 Solutions. r s

MASSACHUSETTS INSTITUTE OF TECHNOLOGY Physics Department. Problem Set 10 Solutions. r s MASSACHUSETTS INSTITUTE OF TECHNOLOGY Physics Depatment Physics 8.033 Decembe 5, 003 Poblem Set 10 Solutions Poblem 1 M s y x test paticle The figue above depicts the geomety of the poblem. The position

More information

Goodness-of-fit for composite hypotheses.

Goodness-of-fit for composite hypotheses. Section 11 Goodness-of-fit fo composite hypotheses. Example. Let us conside a Matlab example. Let us geneate 50 obsevations fom N(1, 2): X=nomnd(1,2,50,1); Then, unning a chi-squaed goodness-of-fit test

More information

A Comparison and Contrast of Some Methods for Sample Quartiles

A Comparison and Contrast of Some Methods for Sample Quartiles A Compaison and Contast of Some Methods fo Sample Quatiles Anwa H. Joade and aja M. Latif King Fahd Univesity of Petoleum & Mineals ABSTACT A emainde epesentation of the sample size n = 4m ( =, 1, 2, 3)

More information

Information Retrieval Advanced IR models. Luca Bondi

Information Retrieval Advanced IR models. Luca Bondi Advanced IR models Luca Bondi Advanced IR models 2 (LSI) Pobabilistic Latent Semantic Analysis (plsa) Vecto Space Model 3 Stating point: Vecto Space Model Documents and queies epesented as vectos in the

More information

Binary Codes with Locality for Multiple Erasures Having Short Block Length

Binary Codes with Locality for Multiple Erasures Having Short Block Length Binay Codes with Locality fo Multiple Easues Having Shot Bloc Length S. B. Balaji, K. P. Pasanth and P. Vijay Kuma, Fellow, IEEE Depatment of Electical Communication Engineeing, Indian Institute of Science,

More information