STABILITY RESULTS INVOLVING SURFACE AREA MEASURES OF CONVEX BODIES DANIEL HUG AND ROLF SCHNEIDER

Size: px
Start display at page:

Download "STABILITY RESULTS INVOLVING SURFACE AREA MEASURES OF CONVEX BODIES DANIEL HUG AND ROLF SCHNEIDER"

Transcription

1 STABILITY RESULTS INVOLVING SURFACE AREA MEASURES OF CONVEX BODIES DANIEL HUG AND ROLF SCHNEIDER Abstract We strengthen some known stability results from the Brunn-Minkowski theory an obtain new results of similar types. These results concern pairs of convex boies for which either surface area measures, or counterparts of such measures in the Brunn-Minkowski-Firey theory, or geometrically significant transforms of such measures, are close to each other. MSC 2000: 52A20, 52A40. 1 Introuction In recent ecaes, several of the classical uniqueness theorems for convex boies have been turne into quantitative versions, in the form of stability results. The starting point for the present investigation are uniqueness theorems of Minkowski an Aleksanrov, respectively. Minkowski s theorem, in its later general form, says that a -imensional convex boy is uniquely etermine, up to a translation, by its ( 1)st surface area measure. A theorem of Aleksanrov, inepenently prove by Fenchel an Jessen, states the extension of this result to lower orer surface area measures. Aleksanrov s projection theorem asserts that a -imensional convex boy with a given centre of symmetry is uniquely etermine by the volumes (or intrinsic volumes of a given positive orer) of its ( 1)-imensional orthogonal projections. Also this theorem involves surface area measures, since volumes of ( 1)-imensional orthogonal projections an area measures are relate by the cosine transform, see (6). In the following, we improve some known stability results corresponing to these uniqueness theorems, an we obtain new stability versions of some similar uniqueness assertions. To formulate a stability version of Minkowski s uniqueness theorem, we enote by K (r, R) the set of convex boies in Eucliean space R which contain some ball of raius r > 0 an are containe in some ball of raius R > r. Let K, L K (r, R) be convex boies whose surface area measures S 1 (K, ) an S 1 (L, ) satisfy S 1 (K, ) S 1 (L, ) ɛ. (1) A typical stability version of Minkowski s theorem requires to fin a number α > 0, which epens only on, an a number c > 0 epening only on, r, R such that for ɛ 0, inequality (1) implies δ(k, L + x) cɛ α (2) 1

2 with a suitable x R ; here δ enotes the Hausorff metric. The best result of this type up to now is ue to Diskant [4] (Theorem in [23]); it gives a stability orer α = 1/. Uner the stronger assumption that K, L are of class C 2 + an that S 1 (K, ) S 1 (L, ) ɛσ, (3) where σ enotes the spherical Lebesgue measure, Diskant [6] (with etails in [7]) obtaine (2) with the better exponent α = 1/( 1). Our first result, to be prove in Section 2, will achieve (2) with α = 1/( 1) uner the weaker assumption (1), for a large class of convex boies incluing polytopes an boies of class C 2 +. We also show that in this result the exponent 1/( 1) is optimal. Assumption (1) is essentially equivalent to an assumption on the total variation norm of the ifference of the surface area measures of K an L: (1) implies S 1 (K, ) S 1 (L, ) T V 2ɛ, (4) an (4) implies (1) with ɛ replace by 2ɛ. Following a suggestion of Wolfgang Weil, we replace this assumption on the total variation istance of measures by a more natural one on the Prohorov istance P. The inequality (1) implies P (S 1 (K, ), S 1 (L, )) ɛ, (5) but not conversely. We show in Section 3 that the weaker assumption (5) is still sufficient to obtain the stability estimate (2) for all convex boies K, L K (r, R) with α = 1/. For the Aleksanrov-Fenchel-Jessen theorem on lower orer surface area measures (Corollary in [23]), a stability version was obtaine by Schneier [22] (Theorem in [23]). Lutwak s work on the Brunn-Minkowski-Firey theory, where Minkowski sums of convex boies are replace by Firey s p-sums, contains also a generalization of the Aleksanrov-Fenchel-Jessen theorem (Corollary (2.3) in [19]). In Section 4 we will give a stability result for this extene theorem. As corollaries of the proof, we obtain stability versions of two inequalities of Lutwak [19]. For a convex boy K R an a unit vector u S 1, we enote by K u the image of K uner orthogonal projection onto u, the hyperplane through 0 orthogonal to u. We write V 1 ( ) for the volume in ( 1)-imensional hyperplanes. Then V 1 (K u ) = 1 u, v S 1 (K, v), (6) 2 S 1 where, is the scalar prouct of R. Thus, the projection function u V 1 (K u ) of K is, up to a constant factor, the cosine transform of the surface area measure S 1 (K, ). A special case of Aleksanrov s projection theorem (e.g., Theorem in [10]) says that two -imensional centrally symmetric convex boies with the same projection function iffer only by a translation. Stability 2

3 versions of this uniqueness theorem are ue to Campi [3] (for = 3) an to Bourgain & Linenstrauss [2]. In Section 5 we use the metho of Bourgain an Linenstrauss to obtain further stability results of a similar nature. One of these results concerns the sine transform u 1 u, v 2 S 1 (K, v) S 1 of the surface area measure, which also has geometric significance. Then we obtain some stability results for convex boies which are not necessarily centrally symmetric. They refer to various integral transforms, appearing in work of Anikonov & Stepanov [1], Gooey & Weil [12], Schneier [24]. 2 Stability for Minkowski s theorem We work in -imensional real vector space R ( 3), equippe with the stanar Eucliean structure. The set of convex boies (non-empty compact convex sets) in R is enote by K. For notions from the theory of convex boies which are not explaine here, we refer to [23]. Apart from replacing E n by R, we use the terminology of that book. Let K K be a convex boy, an let S 1 (K, ) be its surface area measure of orer 1. It is a finite Borel measure on the unit sphere S 1. By Lebesgue s ecomposition theorem, it can be ecompose, with respect to the ( 1)-imensional Hausorff measure H 1, into an absolutely continuous part S 1 a (K, ) an a singular part Ss 1 (K, ). The latter can be ecompose further, by efining S 1(K, c ) := S 1 (K, {u})δ u, (7) u S 1 where δ u enotes the Dirac measure (unit point mass) at u, an S n 1(K, ) := S s 1(K, ) S c 1(K, ). Clearly, in (7) at most countably many summans are non-zero. Moreover, S 1 (K, {u}) = H 1 (F (K, u)) for u S 1, where F (K, u) is the support set of K with outer normal vector u. Thus, we have the ecomposition S 1 (K, ) = S a 1(K, ) + S c 1(K, ) + S n 1(K, ) (8) of the surface area measure S 1 (K, ) into an absolutely continuous measure S 1 a (K, ), a component Sc 1 (K, ) which is an at most countable sum of point masses, an a singular component S 1 n (K, ) without point masses. 2.1 Theorem. Let 0 < r < R. There exists a number c, which epens only on, r, R, with the following property. If K, L K (r, R) are convex boies satisfying 3

4 S 1 n (K, ) = 0, Sn 1 (L, ) = 0 an the assumption S 1 (K, ) S 1 (L, ) ɛ (9) for some ɛ 0, then for a suitable vector x R. δ(k, L + x) cɛ 1 1 The conition S 1 n (K, ) = 0 is fulfille, for example, if K is a polytope, or if the surface area measure of K is absolutely continuous. The latter is true, in particular, if the support function of K is of class C 2. The exponent 1/( 1) in Theorem 2.1 is optimal, at least for the class of polytopes. This can be seen by choosing for K a unit cube an for L the polytope which is obtaine from K by cutting off a vertex of K in such a way that the section plane meets K in a regular ( 1)-simplex of ege length ɛ 1/( 1). On the other han, uner the stronger assumption that L is a ball an that (3) hols, Theorem 3.4 in [17] achieves (2) with α = 1. The subsequent proof of Theorem 2.1 is a refinement of the approach of Diskant [6], [7] an makes also use of Diskant [5]. The proof oes not require the full conition S 1 n (K, ) = 0 = Sn 1 (L, ), but only its consequence S n 1((1 t)k + tl, ) max{s n 1(K, ), S n 1(L, )} for 0 t 1; (10) hence we will work uner this assumption. As usual, we write V 1 (K, L) := V (K[ 1], L) for K, L K, where V enotes the mixe volume, thus V 1 (K, L) = 1 h(l, u)s 1 (K, u). S 1 Here h(l, ) is the support function of L. We also write V ( ) for the volume functional in R. We assume that K, L K (r, R) are convex boies satisfying (9) an (10). For t [0, 1] we set H t := (1 t)k + tl, then H t K (r, R). In the following, c 1, c 2,... enote positive constants which epen only on, r, R. The proof is ivie into four steps. Step I. First we show that V (H t ) V (K) c 1 ɛ, V (H t ) V (L) c 1 ɛ (11) 4

5 for t [0, 1]. By Lemma in [23], the estimates V (L) V 1 (K, L) c 2 ɛ, (12) 0 V 1 (K, L) V (K) 1 V (L) 1 c3 ɛ (13) an the corresponing ones with K an L interchange follow from the assumptions on K an L. From V 1 (K, L) V (K) 1 V (L) an (12) we euce that hence ( V (L) 1 + c ) 2ɛ = (V (L) + c 2 ɛ) V (K) 1 V (L), V (L) By symmetry, we infer that V (K) V (L) (1 + c 4 ɛ) 1 V (L) + c 5 ɛ. V (K) V (L) c 5 ɛ. Since K, L K (r, R), this implies V (K) 1/ V (L) 1/ c6 ɛ. (14) The function φ efine by φ(t) := V (H t ) 1/ (1 t)v (K) 1/ tv (L) 1/, t [0, 1], is concave an satisfies φ(0) = φ(1) = 0, hence Using (13) an K K (r, R) we get φ (0) φ(t) 0 for t [0, 1]. (15) φ (0) = V 1(K, L) V (K) 1 V (L) 1 V (K) 1 c 7 ɛ. (16) Hence, by (14), (15) an (16) we obtain V (H t ) 1/ V (K) 1/ V (L) 1/ V (K) 1/ + c7 ɛ c 8 ɛ. Since H t, K K (r, R), this implies the first inequality of (11), an the secon follows by symmetry. Step II. Next we show an analogue of (11) for surface areas, namely V1 (H t, B ) V 1 (K, B ) c9 ɛ, V1 (H t, B ) V 1 (L, B ) c9 ɛ (17) 5

6 for t [0, 1]; here B is the unit ball. If the support function h(m, ) of the convex boy M K is secon orer ifferentiable at u S 1 (which hols for H 1 almost all u S 1 ), then the eigenvalues of the secon orer ifferential 2 h(m, ) u at u are the principal raii of curvature of M at the point with outer normal vector u. Their prouct is enote by D 1 h(m, u), thus D 1 h(m, u) = et( 2 h(m, u) u ). For all u S 1 with the property that h(k, ) an h(l, ) are secon orer ifferentiable at u, an thus for H 1 almost all u S 1, we set m(u) := min{d 1 h(k, u), D 1 h(l, u)}. Minkowski s eterminant inequality states that hence D 1 h(h t, u) 1 1 (1 t)d 1 h(k, u) td 1 h(l, u) 1 1, D 1 h(h t, u) m(u) (18) for all t [0, 1]. Let ω 1 enote the measurable set of all u S 1 such that h(k, ) an h(l, ) are secon orer ifferentiable at u an D 1 h(k, u) > m(u). Then m(u) = D 1 h(l, u) for u ω 1 an 0 (D 1 h(k, u) m(u))h 1 (u) S 1 = (D 1 h(k, u) D 1 h(l, u))h 1 (u) ω 1 = S 1 (K, ω 1 ) S 1 (L, ω 1 ). Here we have use the fact that if M K an ω is the set of all u S 1 such that h(m, ) is secon orer ifferentiable at u, then the restriction of the measure S 1 (M, ) to ω is absolutely continuous with respect to H 1. This can be verifie on the basis of [15], [16] (an with the terminology use there): Choose any normal bounary point x τ(m, ω) of M, that is, x = τ M (u) for a uniquely etermine u ω. Then by an inspection of the proofs of Lemma 3.4 an Lemma 3.1 in [15], we obtain k i (x, u) > 0 for i {1,..., 1}; moreover, by Lemma 3.1 in [15], we fin k i (x) = k i (x, u) for i {1,..., 1}. This shows that H 1 (M, x) > 0 for H 1 almost all x τ(m, ω). The assertion is then implie by Theorem 3.7 in [16]. Now it follows from (9) that 0 (D 1 h(k, u) m(u))h 1 (u) ɛ. (19) S 1 6

7 an For M K an u S 1, we set f(m, u) := V 1 (F (M, u)) m(u) := min{f(k, u), f(l, u)}. The aitivity of support sets ([23], Theorem 1.7.5(c)), together with the Brunn- Minkowski theorem, implies that an therefore f(h t, u) 1 1 (1 t)f(k, u) tf(l, u) 1 1 f(h t, u) m(u) (20) for u S 1 an all t [0, 1]. Let ω 2 enote the set of all u S 1 such that f(k, u) > m(u). Hence, ω 2 is at most countable, an for u ω 2 we have m(u) = f(l, u) an f(k, u) m(u) = S 1 (K, {u}) S 1 (L, {u}), thus 0 (f(k, u) m(u)) = (f(k, u) m(u)) u S 1 u ω 2 = S 1 (K, ω 2 ) S 1 (L, ω 2 ). Therefore, (9) implies 0 u S 1 (f(k, u) m(u)) ɛ. (21) For convex boies M, H K, we now make use of the ecomposition V 1 (M, H) = 1 h(h, u) [ S 1(M, a u) + S 1(M, c u) + S 1(M, n u) ] S 1 = V a 1 (M, H) + V c 1 (M, H) + V n 1 (M, H) with V1 a (M, H) := 1 h(h, u)d 1 h(m, u)h 1 (u), V c 1 (M, H) := 1 h(h, u)f(m, u), V1 n (M, H) := 1 h(h, u)s n 1(M, u). 7

8 Here, as below, we write instea of if the integration is extene over S 1. S 1 Similarly, means u S 1, where the summation effectively extens only over countably many summans. We estimate the expression I := V (H t ) V1 n (K, H t ) 1 h(h t, u)m(u)h 1 (u) 1 h(ht, u)m(u) from both sies. Without loss of generality, we assume that rb K, L. Then the support function of H t satisfies r h(h t, ) 2R. First we insert in I the expression V n 1 (K, H t ) = V 1 (K, H t ) + V a 1 (K, H t ) + V c 1 (K, H t ) an use H t = (1 t)k + tl together with the estimates (11), (19), (21) to obtain I = V (H t ) 1 h(h t, u)s 1 (K, u) + 1 h(h t, u)(d 1 h(k, u) m(u))h 1 (u) + 1 h(ht, u)(f(k, u) m(u))h 1 (u) (1 t)(v (H t ) V (K)) + t(v (H t ) V 1 (K, L)) + 2R [ (D 1 h(k, u) m(u))h 1 (u) + ] (f(k, u) m(u)) t(v (H t ) V 1 (K, L)) + c 10 ɛ. (22) Next, we insert in I the ecomposition V (H t ) = V a 1 (H t, H t ) + V c 1 (H t, H t ) + V n 1 (H t, H t ) an use the fact that S 1 n (H t, ) S 1 n (K, ) is, by (10), a positive measure. Together with (18) an (20), this gives I = 1 h(h t, u) ( S n 1(H t, u) S 1(K, n u) ) + 1 h(h t, u) (D 1 h(h t, u) m(u)) H 1 (u) + 1 h(ht, u)(f(h t, u) m(u)) r { S n 1 ( Ht, S 1) S n 1(K, S 1 ) 8

9 + (D 1 h(h t, u) m(u))h 1 (u) + (f(h t, u) m(u)) } r { S n 1 (H t, S 1 ) S n 1(K, S 1 ) + (D 1 h(h t, u) D 1 h(k, u))h 1 (u) + (f(h t, u) f(k, u)) } = r(v 1 (H t, B ) V 1 (K, B )). (23) Combining (22) an (23), we fin By (11) an (12), an thus V 1 (H t, B ) V 1 (K, B ) t r (V (H t) V 1 (K, L)) + c 11 ɛ. V (H t ) V 1 (K, L) V (H t ) V (L) + V (L) V 1 (K, L) c 12 ɛ V 1 (H t, B ) V 1 (K, B ) c 13 ɛ, t [0, 1]. (24) On the other han, from (18), (20), (10), (19), (21) we euce V 1 (H t, B ) = V1 a (H t, B ) + V1 c (H t, B ) + V1 n (H t, B ) 1 m(u)h 1 (u) m(u) + Sn 1(H t, S 1 ) = 1 D 1 h(k, u)h 1 (u) f(k, u) + Sn 1(K, S 1 ) 1 (D 1 h(k, u) m(u))h 1 (u) 1 1 ( ( (f(k, u) m(u)) + S n 1 Ht, S 1) S 1(K, n S 1 ) ) V 1 (K, B ) 2 ɛ. (25) Now (24) an (25) yiel the first estimate of (17), an the secon follows by interchanging K an L. The next step provies corresponing estimates for projection volumes. Step III. If v S 1 an t [0, 1], then V 1 (H v t ) V 1 (K v ) c 14 ɛ, V 1 (H v t ) V 1 (L v ) c 14 ɛ. (26) 9

10 For the proof, we efine J := V 1 (Ht v ) 1 u, v S n 2 1(K, u) 1 u, v m(u)h 1 (u) 1 u, v m(u). 2 2 Since J = 1 2 u, v (D 1 h(h t, u) m(u))h 1 (u) + 1 u, v (f(ht, u) m(u)) u, v ( S n 2 1(H t, u) S 1(K, n u) ) 0, we can euce that V 1 (H v t ) V 1 (K v ) 1 2 u, v (D 1 h(k, u) m(u))h 1 (u) 1 u, v (f(k, u) m(u)) 2 ɛ. (27) On the other han, by (17), (19), (21) V 1 (Ht v ) V 1 (K v ) J = 1 2 u, v ( S 1(H n t, u) S 1(K, n u) ) u, v (D 1 h(h t, u) m(u)) H 1 (u) + 1 u, v (f(ht, u) m(u)) 2 S 1(H n t, S 1 ) S 1(K, n S 1 ) + (D 1 h(h t, u) m(u))h 1 (u) + (f(h t, u) m(u)) = (V 1 (H t, B ) V 1 (K, B )) 10

11 + (D 1 h(k, u) m(u))h 1 (u) + (f(k, u) m(u)) c 15 ɛ. (28) The estimates (27) an (28) yiel the first estimate in (26), an the secon estimate follows by symmetry. Step IV. The rest of the proof now follows from the work of Diskant [5], [7]. For given v S 1, the function efine by φ v (K, L, t) := V 1 (H v t ) 1 1 (1 t)v 1 (K v ) 1 1 tv 1 (L v ) 1 1 for t [0, 1] can be estimate, in view of (26), by ] φ v (K, L, t) = (1 t) [V 1 (Ht v ) 1 1 V 1 (K v ) 1 1 +t [V 1 (Ht v ) 1 1 V 1 (L v ) 1] c 16 ɛ. If λ > 0 is such that V 1 (λk v ) = V 1 (L v ), one obtains φ v (λk, L, t) c 17 ɛ. Now the main theorem of [5] shows that there exist ɛ 0 > 0 an c 18, epening only on, r, R, such that δ(λk v, L v + x(v)) c 18 ɛ 1 1 for some x(v) v, if ɛ ɛ 0, an therefore δ(k v, L v + x(v)) c 19 ɛ 1 1. For ɛ > ɛ 0, the same inequality hols if the constant c 19 is ajuste. Thus the assertion of Theorem 2.1 is implie by Theorem in [10]. 3 Stability an Prohorov metric For a set A S 1 an for ɛ > 0, let A ɛ := {y S 1 : x y < ɛ for some x A}, where is the Eucliean norm. For finite Borel measures µ, ν on S 1, let P (µ, ν) := inf { ɛ > 0 : µ(a) ν(a ɛ ) + ɛ an ν(a) µ(a ɛ ) + ɛ for all Borel sets A S 1}. 11

12 This efines the Prohorov metric P, which metrizes the weak topology (e.g., see [8], Section 11.3, in the case of probability measures). The following theorem strengthens Diskant s stability theorem (Theorem in [23]), replacing the assumption (1) by the weaker assumption (5). 3.1 Theorem. Let 0 < r < R. There exists a number c, epening only on, r, R, such that, for K, L K (r, R), for some x R. δ(k, L + x) c P (S 1 (K, ), S 1 (L, )) 1/ Proof. In the following, the positive constants c 1, c 2,... epen only on, r, R. Let K, L K (r, R). We set µ := S 1 (K, ), ν := S 1 (L, ), µ 1 := µ(s 1 ), ν 1 := ν(s 1 ), ɛ := P (µ, ν). Then an hence µ 1 1 ν 1 c 1ɛ, µ 1 ν 1 ɛ, µ 1, ν 1 1 c 1, ν 1 1 µ 1 c 1ɛ. For any Borel set A S 1 we euce that µ(a) ν ( 1 ν(aɛ ) + ɛ ) ( ν(aɛ ) (1 + c 1 ɛ) + ɛ ) ν(a ɛ) + c 2 ɛ. µ 1 µ 1 ν 1 ν 1 ν 1 ν 1 ν 1 By symmetry, we fin For a function f : S 1 R we set P ( µ µ 1, ν ν 1 ) c 2 ɛ. f L := sup x y f(x) f(y), (29) x y f := sup f(x), f BL := f L + f. (30) x It follows from the proof of Corollary in [8] that ( µ f ν ) ( µ 2 f BL P, ν ). µ 1 ν 1 µ 1 ν 1 Thus, for any function f : S 1 R with f BL 1 we get [ ( f(µ ν) µ µ 1 f ν ) + 1 µ 1 ν 1 1 ν 1 µ 1 ] f ν µ 1 (2c 2 ɛ + c 3 ɛ) = c 4 ɛ. 12

13 We may assume that K RB, then h(k, ) BL 2R (cf. [23], Lemma ). Therefore, V (K) V 1 (L, K) = 1 h(k, u)(µ ν)(u) 2R S c 4ɛ = c 5 ɛ, 1 similarly V (L) V 1 (K, L) c 5 ɛ. These estimates correspon to the inequalities (7.2.6) in [23], an the proof can now be complete as the proof of Lemma an of Theorem in [23]. Note that the latter proof gives δ(k, L+x) cɛ 1/ if ɛ is smaller than a certain positive constant ɛ 1 epening only on, r, R; if ɛ ɛ 1, then the same inequality is achieve by a suitable choice of c. 4 Stability results in the Brunn-Minkowski-Firey theory A basic notion of the Brunn-Minkowski theory is the vector aition of convex boies, which correspons to the aition of support functions, h(k + L, ) = h(k, ) + h(l, ). For p 1, a p-sum of convex boies K, L K 0 (the set of convex boies in R with 0 as interior point) can be efine by h(k + p L, ) := [h(k, ) p + h(l, ) p ] 1/p, since the right-han sie is again a support function. Such p-means of convex boies were introuce by Firey [9]. Lutwak [19], [20] has extene large parts of the Brunn-Minkowski theory to this more general combination of convex boies. In this Brunn-Minkowski-Firey theory, as it is now calle, the role of the classical surface area measures S m (K, ), m = 0,..., 1 (see, e.g., Section 4.2 of [23]) is playe by measures S p,i (K, ) on the sphere S 1 (i = 0,..., 1). The measure S p,i (K, ) is absolutely continuous with respect to S 1 i (K, ) an has a Raon- Nikoym erivative given by S p,i (K, ) S 1 i (K, ) = h(k, )1 p. (Note that in [19] the measure S 1,i (K, ) is, unfortunately, enote by S i (K, ) an not by S 1 i (K, ), as usual.) Lutwak s theory contains an analogue of the Aleksanrov-Fenchel-Jessen theorem, Corollary (2.3) of [19]: Suppose K, L K 0 an 0 i <. If i p > 1 an S p,i (K, ) = S p,i (L, ), then K = L. In the following, we obtain a stability version of this result. Again, we use an assumption on the Prohorov istance of two measures, which is weaker than the corresponing assumption for the total variation istance of the measures. 13

14 By K 0(r, R) we enote the set of convex boies K R which satisfy rb K RB, where 0 < r < R. 4.1 Theorem. Let p > 1 an 0 < r < R. Suppose that K, L K 0(r, R), i {0,..., 1}, i p, an with some ɛ 0. Then P (S p,i (K, ), S p,i (L, )) ɛ (31) δ(k, L) cɛ q/2 with q = where the constant c epens only on, p, r, R. 1 ( + 1)2 i 2, Proof. In the following, c 1, c 2,... enote positive constants which epen only on, p, r, R. In the subsequent estimations where such constants occur, we very often tacitly use the facts that rb K, L RB, an that mixe volumes are monotone in each argument. With K an L as in the theorem, we use the notations (all integrations are over the sphere S 1 ) W i (K) = 1 h(k, u)s 1 i (K, u), W i (K, L) = 1 h(l, u)s 1 i (K, u) = V (K[ 1 i], L[1], B [i]), W p,i (K, L) = 1 h(l, u) p S p,i (K, u) = 1 h(l, u) p h(k, u) 1 p S 1 i (K, u). As in [23], p. 398, we write, for some fixe i {0,..., 1} an for k {0,..., i}, V (k) := V (K[ i k], L[k], B [i]), thus W i (K) = V (0), W i (K, L) = V (1), W i (L) = V ( i). With these notations, Lutwak s [19] inequality (IIp) (p. 132; see also Theorem 1.2 in [19]) reas Interchanging K an L, we get W p,i (K, L) i V i p (0) V p ( i). (32) W p,i (L, K) i V i p ( i) V p (0). (33) 14

15 Another inequality prove by Lutwak [19] (p. 137) states that W p,i (K, L) W i (K, L) p W i (K) 1 p = V p (1) V 1 p (0). (34) Using (31), we can estimate as in Section 3 an obtain V(0) W p,i (L, K) = 1 h(k, u) p [S p,i (K, u) S p,i (L, u)] c 1 h(k, ) p BL ɛ, hence an similarly We write V (0) W p,i (L, K) c 2 ɛ (35) W p,i (K, L) V ( i) c 3 ɛ. (36) W p,i (K, L) V i p i = ( V( i) + V (0) ( V( i) V (0) (0) V p i ( i) ) p i ( V i p i ( i) ) V p i (0) W p,i (L, K) ) p i [Wp,i (L, K) V (0) ] + [ Wp,i (K, L) V ( i) ]. By (33), the first term on the right is not positive, hence (35) an (36) give W p,i (K, L) V i p i (0) V p i ( i) c 4ɛ. (37) Now we assume that i {0,..., 2}. We write (37) in the form [ ] [ ] W p,i (K, L) V p (1) V 1 p (0) + V p p( 1 i) (1) V i (0) V p i ( i) V 1 p (0) c 4 ɛ. Here both brackets are nonnegative, the first by (34), an the secon by the Aleksanrov-Fenchel inequalities. We euce that an W p,i (K, L) V p (1) V 1 p (0) c 4 ɛ (38) V p p( 1 i) (1) V i Interchanging K an L in (39) gives V p (0) V p i ( i) + c 5ɛ. (39) p( 1 i) ( 1 i) V i ( i) V p i (0) + c 5 ɛ. (40) 15

16 Multiplication of (39) an (40) yiels V (1) V ( 1 i) V (0) V ( i) + c 6 ɛ. (41) We are now in the same situation as in the proof of Theorem in [23]: the inequality there before (7.2.12) is precisely (41), with m replace by i an c 2 replace by c 6. Hence, the subsequent arguments in [23] (see the Appenix of the present paper) lea to the conclusion that δ(k, L) c 7 ɛ q (42) (see also the hint at the en of the proof of Theorem 3.1). Here K = [K s(k)]/b(k), where s(k) is the Steiner point an b(k) is the mean with of K. We put λ = b(k)/b(l) an t = s(k) λs(l), then (42) implies δ(k, λl + t) c 8 ɛ q. (43) To erive (34), Höler s inequality was use. In orer to estimate t, we nee a sharper version of that inequality. We use a special case of an inequality by Kober [18], namely w 1 a 1 + w 2 a 2 a w1 1 aw2 2 w(a 1/2 1 a 1/2 2 ) 2 for a 1, a 2 0 an w 1, w 2 > 0 with w 1 + w 2 = 1, where w := min{w 1, w 2 }. Here we put, for p > 1 an a, b > 0, an obtain w 1 = 1 p, w 2 = p 1 p, a 1 = a p b 1 p, a 2 = b a p b 1 p + (p 1)b pa mb 1 p (a p/2 b p/2 ) 2 (44) with m = min{1, p 1}. Write h(m, ) = h M for M K an put I(M) := 1 h M (u) S 1 i (K, u). We apply (44) with a = h L(u) I(L), b = h K(u) I(K), where u S 1, an integrate over all u S 1 with respect to the measure (1/)S 1 i (K, ). The result can be written as W p,i (K, L) V p (1) V 1 p (0) [ ( ) p/2 ( ) ] p/2 2 hl hk 1 c 9 h 1 p K I(L) I(K) S 1 i(k, ). (45) 16

17 The quotient h L /I(L) is invariant uner a ilatation of L, hence on the right-han sie, the boy L can be replace by λl. Therefore, (38) an (45) yiel [ ( ) p/2 ( ) ] p/2 2 hλl hk S 1 i(k, ) c 10ɛ. (46) I(λL) I(K) Since I(M) is invariant uner translations of M, inequality (43) shows that I(λL) I(K) c 11 ɛ q. Using this inequality an the mean value theorem, we can estimate, for u S 1, h λl (u) h K (u) c 12 hλl (u) p/2 h K (u) p/2 ( ) p/2 hλl (u) c 13 I(λL) ( ) p/2 hk (u) I(K) + c 14ɛ q. Together with (46), this yiels an estimate ( c 15 ) 2 h λl h K S 1 i (K, ) h λl h K 2 S 1 i (K, ) c 16 ɛ q. Since h λl+t (u) = h λl (u) + u, t, it follows from (43) that u, t h λl (u) h K (u) + c 8 ɛ q for u S 1. Writing t 1 = t/ t if t 0, we euce that t u, t 1 S 1 i (K, u) c 17 ɛ q/2. Now u, t 1 S 1 i (K, u) c 18, since the integral is, up to a factor epening only on, an intrinsic volume of a projection of K an hence can be estimate from below by a constant epening only on an r. The conclusion is that t c 19 ɛ q/2. (47) To estimate λ, we first euce from (36) an (37) that V ( i) V i p i (0) V p i 17 ( i) c 20ɛ,

18 thus V i p i ( i) V i p i (0) c 21 ɛ. Since we have assume i p 0, this implies From (43), we get hence W i (L) W i (K) c 22 ɛ. K λl + t + c 8 ɛ q B (1 + c 23 ɛ q )λl + t, W i (L) W i (K) + c 22 ɛ W i ((1 + c 23 ɛ q )λl) + c 22 ɛ = [(1 + c 23 ɛ q )λ] i W i (L) + c 22 ɛ. This gives λ 1 c 24 ɛ q. Interchanging the roles of K an L, we similarly obtain λ 1 1 c 25 ɛ q an hence λ 1 c 26 ɛ q. (48) The inequalities (43), (47) an (48) finally give δ(k, L) c 27 ɛ q/2. This completes the proof of Theorem 4.1 in the case where i {0,..., 2}. Finally, we consier the (simpler) case i = 1. As before, we euce that W 1 (L) W 1 (K) c 22 ɛ, an hence by symmetry where λ = b(k)/b(l). Using (37) an (45), we fin that λ 1 c 28 ɛ, (49) [ ( ) p/2 ( ) ] p/2 2 h L h K σ c 29ɛ W 1 (L) W 1 (K) an thus h p/2 λl hp/2 K 2 σ c 30 ɛ. An application of the mean value theorem shows that h λl h K 2 σ c 31 ɛ, hence (49) an Corollary 1 in [25] (see also Lemma in [23]) give δ(k, L) c 32 ɛ The proof of Theorem 4.1 is now complete. 18

19 We remark that the preceing proof also permits us to give stability versions of two inequalities ue to Lutwak [19]. The first of these is his inequality (II p ) (which is (32) above). 4.2 Corollary. Let p > 1 an 0 < r < R. Suppose that K, L K 0(r, R), i {0,..., 1} an W p,i (K, L) W i (K) i p i W i (L) p i ɛ (50) with some ɛ 0. Then there is a constant c epening only on, p, r, R such that δ(k, λl) cɛ q/2, where λ = b(k)/b(l) an q is as in Theorem 4.1. Proof. Assume that i {0,..., 2}. Then the assumption (50) implies that δ(k, L) c 33 ɛ q, by the argument after equation (37). The subsequent argument in the proof of Theorem 4.1, which shows that s(k) λs(l) c 34 ɛ q/2, remains the same, hence δ(k, λl) c 35 ɛ q/2, as state. The case i = 1 can be treate as in the proof of Theorem 4.1. The next result gives a stability version of Lutwak s Corollary (1.3) (using his notations). 4.3 Corollary. Let p > 1, 0 < r < R an ϑ (0, 1). Suppose that K, L K 0(r, R), i {0,..., 1} an W i ((1 ϑ) K + p ϑ L) p i (1 ϑ)wi (K) p i ϑwi (L) p i ɛ (51) with some ɛ 0. Then there is a constant c epening only on, p, r, R such that δ(k, τl) c min{ϑ, 1 ϑ} q/2 ɛ q/2, where τ is a suitable positive constant an q is as in Theorem 4.1. Proof. Put M := (1 ϑ) K + p ϑ L. From the efinitions of p-sums an of the functionals W p,i, we obtain W i (M) = W p,i (M, (1 ϑ) K + p ϑ L) = (1 ϑ)w p,i (M, K) + ϑw p,i (M, L). Since M K0(r, R), we can apply Corollary 4.2 an euce that, with suitable numbers τ 1, τ 2 > 0, ] W i (M) (1 ϑ) [c 36 δ(m, τ 1 K) 2 q + Wi (M) i p i W i (K) p i +ϑ ] [c 37 δ(m, τ 2 L) 2 q + Wi (M) i p i W i (L) p i. 19

20 From this we infer that W i (M) p i (1 ϑ)wi (K) p i ϑwi (L) p i (1 ϑ)c 38 δ(m, τ 1 K) 2 q + ϑc39 δ(m, τ 2 L) 2 q min{ϑ, 1 ϑ}c 40 [δ(m, τ 1 K) + δ(m, τ 2 L)] 2 q min{ϑ, 1 ϑ}c 41 δ(τ 1 K, τ 2 L) 2 q min{ϑ, 1 ϑ}c42 δ(k, τl) 2 q. 5 Stability of inverse integral transforms The starting point of this section is formula (6), V 1 (K u ) = 1 u, v S 1 (K, v), u S 1, 2 S 1 which expresses the projection function u V 1 (K u ) of a convex boy K as the cosine transform of its area measure of orer 1. The stability result of Bourgain & Linenstrauss [2] is a quantitative version of the fact that two -imensional convex boies with the same centre of symmetry must be close if their projection functions are close. Groemer s [14] book contains a etaile presentation of this theorem an its proof (Theorem 5.5.7). In the present section, we use the metho of Bourgain an Linenstrauss to obtain stability estimates for the inversion of further integral transforms of area measures occurring in the geometry of convex boies. These integral transforms are of the following type. Let Φ : [ 1, 1] R be a boune, Borel measurable function. For a finite signe Borel measure µ on S 1, let (T Φ µ)(u) := Φ( u, v )µ(v) for u S 1. (52) S 1 For a boune measurable function f on S 1, the transform T Φ f is efine as T Φ µ for the signe measure µ = fσ, where σ enotes spherical Lebesgue measure. We nee a few facts about spherical harmonics, which can all be foun in [14]. If Y n is a spherical harmonic of egree n on S 1, then with T Φ Y n = a,n (Φ)Y n 1 a,n (Φ) = ω 1 Φ(t)Pn(t)(1 t 2 ) ( 3)/2 t, 1 where P n is the Legenre polynomial of imension an egree n (e.g., [14], Th ). Here ω k = kκ k is the area of the k-imensional unit ball, an κ k is its volume. The numbers a,n (Φ) are calle the multipliers of T Φ. 20

21 For f, g L 2 (S 1 ), the space of square integrable real functions on S 1, a scalar prouct is efine by (f, g) := fg σ, S 1 an the L 2 -norm by f := (f, f). Let {Y nj : j = 1,..., N(, n)} be an orthonormal basis of the real vector space of spherical harmonics of egree n N 0. For f L 2 (S 1 ), the relation f Y n (53) means that Y n = N(,n) j=1 n=0 (f, Y nj )Y nj, an the series in (53) is calle the conense harmonic expansion of f ([14], p. 72). Similarly, for a finite signe measure µ on S 1 we write µ Y n (54) if If (54) hols, then Y n = N(,n) j=1 T Φ µ n=0 ( ) Y nj µ Y nj. S 1 a,n (Φ)Y n. (55) n=0 The following theorem is only a slight extension of the result of Bourgain an Linenstrauss, to general transformations T Φ. For the reaer s convenience, we repeat the essential steps of the proof, in a simplifie form, to inicate where changes are necessary. Recall that the norm BL was efine by (30) an that µ T V enotes the total variation norm of the signe measure µ. 5.1 Theorem. Assume that the multipliers of the transformation T Φ satisfy a,0 (Φ) 0, a,n (Φ) 1 bn β for n N (56) with suitable b, β > 0. Let µ be a finite signe measure on S 1, an let F : S 1 R be a Lipschitz function. Then for each α (0, 1/(1 + β)) there is a constant c epening only on, Φ, α such that F µ c F BL µ 1 α T V T Φµ α. S 1 21

22 If µ is even an (56) hols for even n, then the same conclusion can be rawn. Proof. We choose b, β (epening on Φ) so that (56) hols. The constants c 1, c 2,... in the following epen only on, Φ, b, β, α an hence only on, Φ, α. It was the iea of Bourgain an Linenstrauss [2] to use the Poisson transform µ τ := 1 1 τ 2 ω S (1 + τ 1 2 2τ u, v ) µ(v), u /2 S 1, for 0 < τ < 1. We have (all integrations are over S 1 ) F τ µ = F µ τ σ an F µ (F F τ ) µ + F τ µ F F τ µ T V + F µ τ σ. (57) For τ 1/4, F F τ 2 +1 ω 1 2 F L (1 τ) log (58) ω 1 τ ([14], Lemma 5.5.8). Moreover, F µ τ σ F µ τ. (59) If (54) is the conense harmonic expansion of µ, then µ τ τ n Y n. n=0 The maximal value of the function g(x) = x β τ x for x > 0 is ( β/e log τ) β, hence n β τ n (1 τ) β for n N. Therefore, (56) gives ( ) β ( ) β β 1 τ e log τ τ n c 1 (1 τ) β a,n (Φ). Together with Parseval s relation, this yiels µ τ 2 = τ 2n Y n 2 c 2 1(1 τ) 2β n=0 22 n=0 ( ) β β e a,n (Φ) 2 Y n 2.

23 Now (55) shows that µ τ c 1 (1 τ) β T Φ µ. (60) From (57), (58), (59), (60) we get F µ c 2 F T Φ µ (1 τ) β 2 + c 3 F L µ T V (1 τ) log 1 τ ] c 4 F BL [ T Φ µ (1 τ) β 2 + µ T V (1 τ) log. 1 τ Since Φ is boune, we have T Φ µ c 5 µ T V. Therefore, we can fin a constant c 6 an a number τ [ 1 4, 1) such that For this τ an for α (0, 1) we get T Φ µ (1 τ) β = c 6 µ T V (1 τ) log 2 1 τ. ) 1 α F µ c 7 F BL (1 τ) (log 1 α(1+β) 2 µ 1 α T V 1 τ T Φµ α. If now α < 1/(1 + β), then we get F µ c 8 F BL µ 1 α T V T Φµ α. If the signe measure µ is even, then the components in (54) satisfy Y n = 0 for o n. Therefore, one can conclue as above. This completes the proof of Theorem 5.1. The geometric applications are of the following type. 5.2 Theorem. Let Φ an β be as in Theorem 5.1, let 0 < r < R. For γ (0, 1/(1 + β)), there is a constant c epening only on, Φ, γ, r, R with the following property. If K, L K (r, R) an then µ := S 1 (K, ) S 1 (L, ), (61) δ(k, L + x) c T Φ µ γ (62) with a suitable vector x R. If K an L are centrally symmetric an (56) hols for even n, then the same conclusion can be rawn. 23

24 Proof. The constants c 1, c 2,... in this proof epen only on, Φ, γ, r, R. We apply Theorem 5.1 with α = γ to the measure µ given by (61) an to F = h K, the support function of K. Without loss of generality, we assume that K RB. Since µ T V = S 1 (K, S 1 ) + S 1 (L, S 1 ) can be estimate from above by a constant epening only on R an an the same is true for h K BL (cf. [23], Lemma ), we get F µ c 1 T Φ µ α. By the geometric meaning of F an µ, this reas an interchanging K an L we get V (K) V 1 (L, K) c 2 T Φ µ α, V (L) V 1 (K, L) c 2 T Φ µ α. By a result of Diskant [4] (compare the remark at the en of the proof of Theorem 3.1), the two inequalities together imply V (K) V 1 (L, K) ɛ, V (L) V 1 (K, L) ɛ δ(k, L + x) c 3 ɛ 1/ for suitable x R, provie that ɛ ɛ 0, where ɛ 0 > 0 is a constant epening only on, r, R. If c 2 T Φ µ α ɛ 0, then we get δ(k, L + x) c 4 T Φ µ α/, an if c 2 T Φ µ α > ɛ 0, the same estimate hols if c 4 is chosen suitably. If K an L are centrally symmetric, then the signe measure µ is even. This completes the proof of Theorem 5.2. The special case of Theorems 5.1 an 5.2 treate by Bourgain an Linenstrauss concerne the cosine transform, where Φ(t) = 1 2 t for t [ 1, 1]. In that case, (56) hols for even n with β = (+2)/2. Hence, for convex boies K, L K (r, R) with the same centre of symmetry an for the ( 1)st projection function V 1 (K, u) = V 1 (K u ), u S 1, one gets δ(k, L) c V 1 (K, ) V 1 (L, ) γ (63) for γ (0, 2/( + 4)). It is natural to ask for similar results for the ith projection function, V i (K, u) = V i (K u ) = 1 u, v S i (K, v), u S 1. 2 S 1 24

25 By a well-known integral geometric formula, the convex boies K, L satisfy V i (K, ) = V i (L, ) if the projections of K an L on an i-imensional subspace always have the same i-imensional volume. For i = 1, a strong stability result was prove by Gooey an Groemer [13]. In two books, the question for corresponing generalizations was pose. Groemer [14], p. 222, writes that at present such stability estimates exist only in the cases i = 1 an i = 1. Garner [10] asks in his Problem 4.7 (p. 157) whether a stability result of the type (63) can be obtaine for 1 < i < 1. Curiously, a positive answer on the basis of publishe results coul have been given at the time when those books were written. In fact, the analytic part of the Bourgain-Linenstrauss [2] proof (just replace µ by µ i := S i (K, ) S i (L, ) in the first part of the proof of Theorem 5.2) gives V (0) V (i) c 2 T Φ µ i α, V (i+1) V (1) c 2 T Φ µ i α for α (0, 2/( + 4)) (if Φ(t) = 1 2 t ), where V (k) := V (K[i + 1 k], L[k], B [ 1 i]). As shown in [23] (Proof of Lemma 7.2.3), the inequalities V (0) V (i) ɛ, V (i+1) V (1) ɛ for K, L K (r, R) an some ɛ > 0 imply 0 V (1) V i/(i+1) (0) V 1/(i+1) (i+1) ( ) R r + 1 ɛ. One can now essentially use the proof of a stability result for the Aleksanrov- Fenchel-Jessen theorem ([23], Theorem 7.2.6) to euce the following. 5.3 Theorem. Let i {2,..., 2} an 0 < r < R, let K, L K (r, R) be convex boies which are centrally symmetric with the same centre. For ( ) 1 γ 0, ( + 1)( + 4)2 i 2, there exists a constant c epening only on, γ, r, R such that δ(k, L) c V i (K, ) V i (L, ) γ. We turn to other integral transforms of type T Φ which have occurre in geometric contexts. The sine transform is the transformation T Φ with Φ(t) = 1 t 2 for t [ 1, 1]. If K K is a convex boy an u S 1, then V ( 1) (K, u) := = 1 2( + 1) V 2 (K (u + tu)) t 25 S 1 1 u, v 2 S 1 (K, v) (64)

26 (see [21], p. 60); here 2V 2 (K ) is the ( 2)-imensional surface area of a ( 1)- imensional convex boy K. Thus the functional V ( 1) (K, ), the integrate surface area of parallel hyperplane sections, is, up to a factor, the sine transform of the surface area measure of K. The sine transform S is connecte with the cosine transform C an the spherical Raon transform R by the relation RC = κ 2 S, which is easily obtaine by a irect calculation. (For convex boies this implies, in view of (64), that Raon transforms of projection functions are connecte with sections; this interplay was stuie in greater generality by Gooey [11].) This relation implies corresponing relations for the multipliers: if then f Y n, n=0 Cf ζ,n Y n, n=0 κ 2 Sf Rf ρ,n ζ,n Y n. n=0 ρ,n Y n, For even n, we have ζ 1,n = O ( n (+2)/2), as remarke above, an ρ 1,n = O ( n ( 2)/2) (as follows from [14], Lemma an (3.4.19)), hence ρ 1,n ζ 1,n = O(n ). Thus, for the function Φ(t) = 1 t 2, assumption (56) hols for even n with β =. This gives the following result. 5.4 Theorem. Let 0 < r < R, let K, L K (r, R) be convex boies with the same centre of symmetry. For γ (0, 1/( + 1)), there exists a constant c epening only on, γ, r, R such that n=0 δ(k, L) c V ( 1) (K, ) V ( 1) (L, ) γ. The results involving the cosine or sine transform are necessarily restricte to centrally symmetric convex boies, since a transform T Φ µ with an even function Φ oes not contain information on the o part of the signe measure µ. We turn now to stability versions of some uniqueness theorems for not necessarily symmetric convex boies. Anikonov an Stepanov [1] have propose to consier, for K K an u S 1, besies the projection volume V 1 (K, u) = V 1 (K u ), also the area S(K, u) of the illuminate portion of K in irection u, that is, S(K, u) := S 1 (K, {v S 1 : u, v 0}). They showe that the combine functional F (K, u) := pv 1 (K, u) + qs(k, u), u S 1, with constants p, q (p, q, 2pκ 1 +qω 0) etermines the convex boy K uniquely up to a translation. They also prove a corresponing stability result in R 3. This, 26

27 however, is rather weak, since it assumes that the ifference F (K, ) F (L, ) is small in a norm that involves erivatives up to orer six. A stronger result can be obtaine with the ai of Theorem 5.2. In fact, we have F (K, ) = T Φ S 1 (K, ) with Φ = pφ 1 +qφ 2, where Φ 1 (t) = 1 2 t an Φ 2 = 1 [0,1]. Now, a,n (Φ 1 ) = 0 for o n, an a,n (Φ 1 ) 1 = O(n (+2)/2 ) for even n. On the other han, a,n (Φ 2 ) = 0 for even n > 0, an a,n (Φ 2 ) 1 = O(n /2 ) for o n (see [14], Lemma an (3.4.20)). It follows that Φ satisfies (56) with β = ( + 2)/2 (note that the assumption 2pκ 1 + qω 0 ensures that a,0 (Φ) 0). Hence, for any two convex boies K, L K (r, R), we have δ(k, L + x) c F (K, ) F (L, ) γ with a suitable vector x R, for γ (0, 2/( + 4)) an with c epening only on, p, q, γ, r, R. The last two transformations to be consiere stem from the part of theoretical stereology or geometric tomography where one is intereste in obtaining information on convex boies from lower imensional sections. The secon mean section boy M 2 (K) of a convex boy K K was introuce by Gooey an Weil [12]. It is efine by h(m 2 (K), ) = h(k E, ) µ 2 (E). E 2 Here E 2 is the affine Grassmannian of two-imensional planes in R an µ 2 is its motion invariant measure, normalize so that µ 2 ({E E 2 : E B }) = κ 2. Thus, M 2 (K) comprises information about the two-imensional sections of K, in integrate form. Gooey an Weil showe that two -imensional convex boies K an L with M 2 (K) = M 2 (L) iffer only by a translation, an they mentione briefly (on p. 429) that a corresponing stability version coul be obtaine. We will make this more explicit. For unit vectors u, v, let α(u, v) [0, π] enote the angle between u an v. Gooey an Weil (loc. cit., Corollary 2) prove that h(m 2 (K) t, u) = κ 2κ 2 ( ) α(u, v) sin α(u, v) S 1 ( K, v) (65) 2 κ S 1 for u S 1, where t is a suitable translation vector. Their proof (cf. Theorem 2) provies no information about this translation, but a relate remark of Gooey [11], p. 165, gives a hint. Denoting by z r+1 (K) the intrinsic (r + 1)st moment vector of K (see [23], p. 304), we have 1 h(m 2 (K) t, u) u σ(u) = z 1 (M 2 (K) t) = z 1 (M 2 (K)) t κ S 1 = κ 2κ 2 ( 2) κ z 1 (K) t. 27

28 But this must be the zero vector, as we see by using (65), Fubini s theorem, the fact that a vector integral of the form S 1 f( u, v ) u σ(u) is invariant uner rotations fixing v an hence is a multiple of v, an that v S S 1 1 ( K, v) = 0. We euce that the boy ( ) 2 κ M 2(K) := M 2 (K) z 1 (K), κ 2 κ 2 which we call the normalize secon mean section boy of K, satisfies h(m 2(K), u) = α(u, v) sin α(u, v) S 1 ( K, v) for u S 1. S 1 Thus, h(m 2(K), ) = T Φ S 1 ( K, ) with Φ(t) = (arccos t) 1 t 2 for t [ 1, 1]. From a computation in [12] (formula (4.10), together with the relation ( ) n + 3 c n (t) = P 3 n(t) between Gegenbauer an Legenre polynomials, see [14], p. 97) it follows that a,n (Φ) = ( c() (n 1)(n + 1) n!γ ( 1 2 (n + )) (n + 2)!Γ ( 1 2 (n + 2)) with a constant c() epening only on. From this we euce that (56) hols with β =. For the resulting stability estimate, we may now use the Hausorff istance also on the right-han sie: For γ (0, ( + 1)), there exists a constant c epening only on, r, R, γ such that, for K, L K (r, R), for a suitable vector x R. δ(k, L + x) cδ(m 2(K), M 2(L)) γ The origin of our last example is an investigation, [24], on the oriente mean normal measure of a stationary stochastic process of convex particles an its etermination from planar sections. There one has reason to consier the function efine by V ( 1) + (K, u) := H 2 ( u K (u + tu)) t for K K an u S 1 ; here H 2 enotes the ( 2)-imensional Hausorff measure an u K is the upper bounary of K in irection u, that is, the set of ) 2, 28

29 all bounary points of K at which there exists an outer unit normal vector v with u, v 0. By formula (17) of [24], V ( 1) + (K, ) = T Φ S 1 (K, ) with Φ(t) = 1 t 2 1 [0,1] (so that T Φ coul be calle the hemispherical sine transform). For even n N, the multipliers of T Φ are essentially those of the sine transform, namely a,n (Φ) = 1 2 a,n(ψ) for Ψ(t) = 1 t 2. Hence, as shown before Theorem 5.4, a,n (Φ) 1 = O(n ) for even n. For o n, the multipliers have not been etermine explicitly, but it has been shown that a,1 (Φ) 0 an, for o n 3, with 1 3 (n 2) f(n) a,n (Φ) = ω 1 ( 1)( + 1) ( + n 4) (n + 1)(n + 3) n + 3 < ( 1) (n 1)/2 f(n) n + 1 ([24], p. 36 together with (20) an (19)). From this, one obtains a,n (Φ) 1 = O(n /2 ) for o n. We euce that (56) hols with β =. Hence, for 0 < r < R an γ (0, 1/( + 1)) there exists a constant c epening only on, γ, r, R such that, for K K (r, R), with a suitable vector x R. 6 Appenix δ(k, L + x) c V ( 1) + (K, ) V ( 1) + (L, ) γ In the proofs of Theorems 4.1 an 5.3 we have referre to the proof of Theorem in [23], which in turn relies on inequality (6.4.9) of [23] (p. 335). The proof of (6.4.9) given there is not complete, as A. Giannopoulos has kinly pointe out. We take this opportunity to correct the error (using the same notations). The proof of (6.4.9) as given is correct if U 12 U 00 U 01 U 02 < 0; observe that Now, for λ 1, λ 2 0 also 0 V (λ 1 K 1 + λ 2 K 2, K 0, C) 2 U 2 01 U 00 U 11 0, U 2 02 U 00 U (66) V (λ 1 K 1 + λ 2 K 2, λ 1 K 1 + λ 2 K 2, C)V (K 0, K 0, C) = λ 2 1(U 2 01 U 00 U 11 ) + λ 2 2(U 2 02 U 00 U 22 ) 2λ 1 λ 2 (U 12 U 00 U 01 U 02 ). If U 12 U 00 U 01 U 02 > 0, we can euce (6.4.9) from this inequality. If U 12 U 00 U 01 U 02 = 0, (6.4.9) hols by (66). 29

30 References [1] Yu.E. Anikonov an V.N. Stepanov, Uniqueness an stability of the solution of a problem of geometry in the large (in Russian). Mat. Sbornik 116 (1981), Engl. transl.: Math. USSR-Sb. 44 (1981), [2] J. Bourgain an J. Linenstrauss, Projection boies. In Geometric Aspects of Functional Analysis (J. Linenstrauss, V.D. Milman, es.) Lecture Notes in Math. 1317, Springer, Berlin, 1988, pp [3] S. Campi, Recovering a centre convex boy from the areas of its shaows: a stability estimate. Ann. Mat. Pura Appl. 151 (1988), [4] V.I. Diskant, Bouns for the iscrepancy between convex boies in terms of the isoperimetric ifference. Siberian Math. J. 13 (1972), [5] V.I. Diskant, Stability of the solution of the Minkowski equation. Siberian Math. J. 14 (1973), [6] V.I. Diskant, On the question of the orer of the stability function in Minkowski s problem (in Russian). Ukrain. Geom. Sb. 22 (1979), [7] V.I. Diskant, Refinements of an isoperimetric inequality an stability theorems in the theory of convex boies (in Russian). Truy Inst. Mat. (Novosibirsk) 14 (1989), Sovrem. Probl. Geom. Analiz, [8] R.M. Duley, Real Analysis an Probability. Wasworth & Brooks/Cole, Pacific Grove, CA, [9] W.J. Firey, p-means of convex boies. Math. Scan. 10 (1962), [10] R.J. Garner, Geometric Tomography. Encyclopeia of Mathematics an its Applications 58, Cambrige University Press, Cambrige, [11] P. Gooey, Raon transforms of projection functions. Math. Proc. Camb. Phil. Soc. 123 (1998), [12] P. Gooey an W. Weil, The etermination of convex boies from the mean of ranom sections. Math. Proc. Camb. Phil. Soc. 112 (1992), [13] P. Gooey an H. Groemer, Stability results for first orer projection boies. Proc. Amer. Math. Soc. 109 (1990), [14] H. Groemer, Geometric Applications of Fourier Series an Spherical Harmonics. Encyclopeia of Mathematics an its Applications 61, Cambrige University Press, Cambrige, [15] D. Hug, Absolute continuity for curvature measures of convex sets I. Math. Nachr. 195 (1998),

31 [16] D. Hug, Absolute continuity for curvature measures of convex sets II. Math. Z. 232 (1999), [17] D. Hug, Absolute continuity for curvature measures of convex sets III. Av. Math., to appear. [18] H. Kober, On the arithmetic an geometric means an on Höler s inequality. Proc. Amer. Math. Soc. 9 (1958), [19] E. Lutwak, The Brunn-Minkowski-Firey theory I: Mixe volumes an the Minkowski problem. J. Differential Geom. 38 (1993), [20] E. Lutwak, The Brunn-Minkowski-Firey theory II: Affine an geominimal surface areas. Av. Math. 118 (1996), [21] R. Schneier, Über eine Integralgleichung in er Theorie er konvexen Körper. Math. Nachr. 44 (1970), [22] R. Schneier, Stability in the Aleksanrov-Fenchel-Jessen theorem. Mathematika 36 (1989), [23] R. Schneier, Convex Boies: the Brunn-Minkowski Theory. Encyclopeia of Mathematics an its Applications 44, Cambrige University Press, Cambrige, [24] R. Schneier, On the mean normal measures of a particle process. Av. Appl. Prob. (SGSA) 33 (2001), [25] R.A. Vitale, L p metrics for compact, convex sets. J. Approx. Theory 45 (1985), Authors aress: Mathematisches Institut Albert-Luwigs-Universität Eckerstr. 1 D Freiburg i.br. Germany aniel.hug@math.uni-freiburg.e rschnei@uni-freiburg.e 31

On the Orlicz-Brunn-Minkowski theory

On the Orlicz-Brunn-Minkowski theory On the Orlicz-Brunn-Minkowski theory C J Zhao 2 3 4 5 6 7 8 9 0 2 3 4 5 Abstract Recently, Garner, Hug an Weil evelope an Orlicz-Brunn- Minkowski theory Following this, in the paper we further consier

More information

Large Cells in Poisson-Delaunay Tessellations

Large Cells in Poisson-Delaunay Tessellations Large Cells in Poisson-Delaunay Tessellations Daniel Hug an Rolf Schneier Mathematisches Institut, Albert-Luwigs-Universität, D-79104 Freiburg i. Br., Germany aniel.hug, rolf.schneier}@math.uni-freiburg.e

More information

LECTURE NOTES ON DVORETZKY S THEOREM

LECTURE NOTES ON DVORETZKY S THEOREM LECTURE NOTES ON DVORETZKY S THEOREM STEVEN HEILMAN Abstract. We present the first half of the paper [S]. In particular, the results below, unless otherwise state, shoul be attribute to G. Schechtman.

More information

Floating Body, Illumination Body, and Polytopal Approximation

Floating Body, Illumination Body, and Polytopal Approximation Convex Geometric Analysis MSRI Publications Volume 34, 998 Floating Boy, Illumination Boy, an Polytopal Approximation CARSTEN SCHÜTT Abstract. Let K be a convex boy inr an K t its floating boies. There

More information

Banach Journal of Mathematical Analysis ISSN: (electronic)

Banach Journal of Mathematical Analysis ISSN: (electronic) Banach J. Math. Anal. 2 (2008), no., 70 77 Banach Journal of Mathematical Analysis ISSN: 735-8787 (electronic) http://www.math-analysis.org WIDTH-INTEGRALS AND AFFINE SURFACE AREA OF CONVEX BODIES WING-SUM

More information

arxiv: v1 [math.mg] 10 Apr 2018

arxiv: v1 [math.mg] 10 Apr 2018 ON THE VOLUME BOUND IN THE DVORETZKY ROGERS LEMMA FERENC FODOR, MÁRTON NASZÓDI, AND TAMÁS ZARNÓCZ arxiv:1804.03444v1 [math.mg] 10 Apr 2018 Abstract. The classical Dvoretzky Rogers lemma provies a eterministic

More information

Lower bounds on Locality Sensitive Hashing

Lower bounds on Locality Sensitive Hashing Lower bouns on Locality Sensitive Hashing Rajeev Motwani Assaf Naor Rina Panigrahy Abstract Given a metric space (X, X ), c 1, r > 0, an p, q [0, 1], a istribution over mappings H : X N is calle a (r,

More information

12.11 Laplace s Equation in Cylindrical and

12.11 Laplace s Equation in Cylindrical and SEC. 2. Laplace s Equation in Cylinrical an Spherical Coorinates. Potential 593 2. Laplace s Equation in Cylinrical an Spherical Coorinates. Potential One of the most important PDEs in physics an engineering

More information

Function Spaces. 1 Hilbert Spaces

Function Spaces. 1 Hilbert Spaces Function Spaces A function space is a set of functions F that has some structure. Often a nonparametric regression function or classifier is chosen to lie in some function space, where the assume structure

More information

Lecture Introduction. 2 Examples of Measure Concentration. 3 The Johnson-Lindenstrauss Lemma. CS-621 Theory Gems November 28, 2012

Lecture Introduction. 2 Examples of Measure Concentration. 3 The Johnson-Lindenstrauss Lemma. CS-621 Theory Gems November 28, 2012 CS-6 Theory Gems November 8, 0 Lecture Lecturer: Alesaner Mąry Scribes: Alhussein Fawzi, Dorina Thanou Introuction Toay, we will briefly iscuss an important technique in probability theory measure concentration

More information

Gaussian polytopes: variances and limit theorems

Gaussian polytopes: variances and limit theorems Gaussian polytopes: variances an limit theorems Daniel Hug an Matthias Reitzner October 9, 4 Abstract The convex hull of n inepenent ranom points in R chosen accoring to the normal istribution is calle

More information

Monotonicity of facet numbers of random convex hulls

Monotonicity of facet numbers of random convex hulls Monotonicity of facet numbers of ranom convex hulls Gilles Bonnet, Julian Grote, Daniel Temesvari, Christoph Thäle, Nicola Turchi an Florian Wespi arxiv:173.31v1 [math.mg] 7 Mar 17 Abstract Let X 1,...,

More information

Calculus of Variations

Calculus of Variations 16.323 Lecture 5 Calculus of Variations Calculus of Variations Most books cover this material well, but Kirk Chapter 4 oes a particularly nice job. x(t) x* x*+ αδx (1) x*- αδx (1) αδx (1) αδx (1) t f t

More information

CHAPTER 1 : DIFFERENTIABLE MANIFOLDS. 1.1 The definition of a differentiable manifold

CHAPTER 1 : DIFFERENTIABLE MANIFOLDS. 1.1 The definition of a differentiable manifold CHAPTER 1 : DIFFERENTIABLE MANIFOLDS 1.1 The efinition of a ifferentiable manifol Let M be a topological space. This means that we have a family Ω of open sets efine on M. These satisfy (1), M Ω (2) the

More information

Sharp Thresholds. Zachary Hamaker. March 15, 2010

Sharp Thresholds. Zachary Hamaker. March 15, 2010 Sharp Threshols Zachary Hamaker March 15, 2010 Abstract The Kolmogorov Zero-One law states that for tail events on infinite-imensional probability spaces, the probability must be either zero or one. Behavior

More information

Lecture XII. where Φ is called the potential function. Let us introduce spherical coordinates defined through the relations

Lecture XII. where Φ is called the potential function. Let us introduce spherical coordinates defined through the relations Lecture XII Abstract We introuce the Laplace equation in spherical coorinates an apply the metho of separation of variables to solve it. This will generate three linear orinary secon orer ifferential equations:

More information

A stability result for mean width of L p -centroid bodies.

A stability result for mean width of L p -centroid bodies. A stability result for mean with of L p -centroi boies. B. Fleury, O. Guéon an G. Paouris Abstract We give a ifferent proof of a recent result of lartag [1] concerning the concentration of the volume of

More information

LARGE TYPICAL CELLS IN POISSON DELAUNAY MOSAICS

LARGE TYPICAL CELLS IN POISSON DELAUNAY MOSAICS LARGE TYPICAL CELLS IN POISSON DELAUNAY MOSAICS DANIEL HUG an ROLF SCHNEIDER Deicate to Tuor Zamfirescu on the occasion of his sixtieth birthay It is prove that the shape of the typical cell of a Poisson

More information

Acute sets in Euclidean spaces

Acute sets in Euclidean spaces Acute sets in Eucliean spaces Viktor Harangi April, 011 Abstract A finite set H in R is calle an acute set if any angle etermine by three points of H is acute. We examine the maximal carinality α() of

More information

Discrete Mathematics

Discrete Mathematics Discrete Mathematics 309 (009) 86 869 Contents lists available at ScienceDirect Discrete Mathematics journal homepage: wwwelseviercom/locate/isc Profile vectors in the lattice of subspaces Dániel Gerbner

More information

The derivative of a function f(x) is another function, defined in terms of a limiting expression: f(x + δx) f(x)

The derivative of a function f(x) is another function, defined in terms of a limiting expression: f(x + δx) f(x) Y. D. Chong (2016) MH2801: Complex Methos for the Sciences 1. Derivatives The erivative of a function f(x) is another function, efine in terms of a limiting expression: f (x) f (x) lim x δx 0 f(x + δx)

More information

NOTES ON EULER-BOOLE SUMMATION (1) f (l 1) (n) f (l 1) (m) + ( 1)k 1 k! B k (y) f (k) (y) dy,

NOTES ON EULER-BOOLE SUMMATION (1) f (l 1) (n) f (l 1) (m) + ( 1)k 1 k! B k (y) f (k) (y) dy, NOTES ON EULER-BOOLE SUMMATION JONATHAN M BORWEIN, NEIL J CALKIN, AND DANTE MANNA Abstract We stuy a connection between Euler-MacLaurin Summation an Boole Summation suggeste in an AMM note from 196, which

More information

The Principle of Least Action

The Principle of Least Action Chapter 7. The Principle of Least Action 7.1 Force Methos vs. Energy Methos We have so far stuie two istinct ways of analyzing physics problems: force methos, basically consisting of the application of

More information

Quantum Mechanics in Three Dimensions

Quantum Mechanics in Three Dimensions Physics 342 Lecture 20 Quantum Mechanics in Three Dimensions Lecture 20 Physics 342 Quantum Mechanics I Monay, March 24th, 2008 We begin our spherical solutions with the simplest possible case zero potential.

More information

Large Poisson-Voronoi Cells and Crofton Cells

Large Poisson-Voronoi Cells and Crofton Cells Large Poisson-Voronoi Cells an Crofton Cells Daniel Hug Matthias Reitzner Rolf Schneier Abstract. It is prove that the shape of the typical cell of a stationary Poisson-Voronoi tessellation in Eucliean

More information

SINGULAR PERTURBATION AND STATIONARY SOLUTIONS OF PARABOLIC EQUATIONS IN GAUSS-SOBOLEV SPACES

SINGULAR PERTURBATION AND STATIONARY SOLUTIONS OF PARABOLIC EQUATIONS IN GAUSS-SOBOLEV SPACES Communications on Stochastic Analysis Vol. 2, No. 2 (28) 289-36 Serials Publications www.serialspublications.com SINGULAR PERTURBATION AND STATIONARY SOLUTIONS OF PARABOLIC EQUATIONS IN GAUSS-SOBOLEV SPACES

More information

6 General properties of an autonomous system of two first order ODE

6 General properties of an autonomous system of two first order ODE 6 General properties of an autonomous system of two first orer ODE Here we embark on stuying the autonomous system of two first orer ifferential equations of the form ẋ 1 = f 1 (, x 2 ), ẋ 2 = f 2 (, x

More information

1. Aufgabenblatt zur Vorlesung Probability Theory

1. Aufgabenblatt zur Vorlesung Probability Theory 24.10.17 1. Aufgabenblatt zur Vorlesung By (Ω, A, P ) we always enote the unerlying probability space, unless state otherwise. 1. Let r > 0, an efine f(x) = 1 [0, [ (x) exp( r x), x R. a) Show that p f

More information

Lectures - Week 10 Introduction to Ordinary Differential Equations (ODES) First Order Linear ODEs

Lectures - Week 10 Introduction to Ordinary Differential Equations (ODES) First Order Linear ODEs Lectures - Week 10 Introuction to Orinary Differential Equations (ODES) First Orer Linear ODEs When stuying ODEs we are consiering functions of one inepenent variable, e.g., f(x), where x is the inepenent

More information

PDE Notes, Lecture #11

PDE Notes, Lecture #11 PDE Notes, Lecture # from Professor Jalal Shatah s Lectures Febuary 9th, 2009 Sobolev Spaces Recall that for u L loc we can efine the weak erivative Du by Du, φ := udφ φ C0 If v L loc such that Du, φ =

More information

Star bodies with completely symmetric sections

Star bodies with completely symmetric sections Star bodies with completely symmetric sections Sergii Myroshnychenko, Dmitry Ryabogin, and Christos Saroglou Abstract We say that a star body is completely symmetric if it has centroid at the origin and

More information

A sharp Rogers Shephard type inequality for Orlicz-difference body of planar convex bodies

A sharp Rogers Shephard type inequality for Orlicz-difference body of planar convex bodies Proc. Indian Acad. Sci. (Math. Sci. Vol. 124, No. 4, November 2014, pp. 573 580. c Indian Academy of Sciences A sharp Rogers Shephard type inequality for Orlicz-difference body of planar convex bodies

More information

Separation of Variables

Separation of Variables Physics 342 Lecture 1 Separation of Variables Lecture 1 Physics 342 Quantum Mechanics I Monay, January 25th, 2010 There are three basic mathematical tools we nee, an then we can begin working on the physical

More information

Deviation Measures and Normals of Convex Bodies

Deviation Measures and Normals of Convex Bodies Beiträge zur Algebra und Geometrie Contributions to Algebra Geometry Volume 45 (2004), No. 1, 155-167. Deviation Measures Normals of Convex Bodies Dedicated to Professor August Florian on the occasion

More information

ALGEBRAIC AND ANALYTIC PROPERTIES OF ARITHMETIC FUNCTIONS

ALGEBRAIC AND ANALYTIC PROPERTIES OF ARITHMETIC FUNCTIONS ALGEBRAIC AND ANALYTIC PROPERTIES OF ARITHMETIC FUNCTIONS MARK SCHACHNER Abstract. When consiere as an algebraic space, the set of arithmetic functions equippe with the operations of pointwise aition an

More information

Tractability results for weighted Banach spaces of smooth functions

Tractability results for weighted Banach spaces of smooth functions Tractability results for weighte Banach spaces of smooth functions Markus Weimar Mathematisches Institut, Universität Jena Ernst-Abbe-Platz 2, 07740 Jena, Germany email: markus.weimar@uni-jena.e March

More information

L p -Width-Integrals and Affine Surface Areas

L p -Width-Integrals and Affine Surface Areas LIBERTAS MATHEMATICA, vol XXX (2010) L p -Width-Integrals and Affine Surface Areas Chang-jian ZHAO and Mihály BENCZE Abstract. The main purposes of this paper are to establish some new Brunn- Minkowski

More information

NAKAJIMA S PROBLEM: CONVEX BODIES OF CONSTANT WIDTH AND CONSTANT BRIGHTNESS

NAKAJIMA S PROBLEM: CONVEX BODIES OF CONSTANT WIDTH AND CONSTANT BRIGHTNESS NAKAJIMA S PROBLEM: CONVEX BODIES OF CONSTANT WIDTH AND CONSTANT BRIGHTNESS RALPH HOWARD AND DANIEL HUG Dedicated to Rolf Schneider on the occasion of his 65th birthday ABSTRACT. For a convex body K R

More information

An extension of Alexandrov s theorem on second derivatives of convex functions

An extension of Alexandrov s theorem on second derivatives of convex functions Avances in Mathematics 228 (211 2258 2267 www.elsevier.com/locate/aim An extension of Alexanrov s theorem on secon erivatives of convex functions Joseph H.G. Fu 1 Department of Mathematics, University

More information

A new proof of the sharpness of the phase transition for Bernoulli percolation on Z d

A new proof of the sharpness of the phase transition for Bernoulli percolation on Z d A new proof of the sharpness of the phase transition for Bernoulli percolation on Z Hugo Duminil-Copin an Vincent Tassion October 8, 205 Abstract We provie a new proof of the sharpness of the phase transition

More information

Lower Bounds for the Smoothed Number of Pareto optimal Solutions

Lower Bounds for the Smoothed Number of Pareto optimal Solutions Lower Bouns for the Smoothe Number of Pareto optimal Solutions Tobias Brunsch an Heiko Röglin Department of Computer Science, University of Bonn, Germany brunsch@cs.uni-bonn.e, heiko@roeglin.org Abstract.

More information

DECOMPOSITION OF POLYNOMIALS AND APPROXIMATE ROOTS

DECOMPOSITION OF POLYNOMIALS AND APPROXIMATE ROOTS DECOMPOSITION OF POLYNOMIALS AND APPROXIMATE ROOTS ARNAUD BODIN Abstract. We state a kin of Eucliian ivision theorem: given a polynomial P (x) an a ivisor of the egree of P, there exist polynomials h(x),

More information

Sturm-Liouville Theory

Sturm-Liouville Theory LECTURE 5 Sturm-Liouville Theory In the three preceing lectures I emonstrate the utility of Fourier series in solving PDE/BVPs. As we ll now see, Fourier series are just the tip of the iceberg of the theory

More information

The Three-dimensional Schödinger Equation

The Three-dimensional Schödinger Equation The Three-imensional Schöinger Equation R. L. Herman November 7, 016 Schröinger Equation in Spherical Coorinates We seek to solve the Schröinger equation with spherical symmetry using the metho of separation

More information

ON SUCCESSIVE RADII AND p-sums OF CONVEX BODIES

ON SUCCESSIVE RADII AND p-sums OF CONVEX BODIES ON SUCCESSIVE RADII AND p-sums OF CONVEX BODIES BERNARDO GONZÁLEZ AND MARÍA A. HERNÁNDEZ CIFRE Abstract. We study the behavior of the so called successive inner and outer radii with respect to the p-sums

More information

INRADII OF SIMPLICES

INRADII OF SIMPLICES INRADII OF SIMPLICES ULRICH BETKE, MARTIN HENK, AND LYDIA TSINTSIFA Abstract. We stuy the following generalization of the inraius: For a convex boy K in the -imensional Eucliean space an a linear k-plane

More information

SOME RESULTS ON THE GEOMETRY OF MINKOWSKI PLANE. Bing Ye Wu

SOME RESULTS ON THE GEOMETRY OF MINKOWSKI PLANE. Bing Ye Wu ARCHIVUM MATHEMATICUM (BRNO Tomus 46 (21, 177 184 SOME RESULTS ON THE GEOMETRY OF MINKOWSKI PLANE Bing Ye Wu Abstract. In this paper we stuy the geometry of Minkowski plane an obtain some results. We focus

More information

Agmon Kolmogorov Inequalities on l 2 (Z d )

Agmon Kolmogorov Inequalities on l 2 (Z d ) Journal of Mathematics Research; Vol. 6, No. ; 04 ISSN 96-9795 E-ISSN 96-9809 Publishe by Canaian Center of Science an Eucation Agmon Kolmogorov Inequalities on l (Z ) Arman Sahovic Mathematics Department,

More information

Implicit Differentiation

Implicit Differentiation Implicit Differentiation Thus far, the functions we have been concerne with have been efine explicitly. A function is efine explicitly if the output is given irectly in terms of the input. For instance,

More information

Introduction to the Vlasov-Poisson system

Introduction to the Vlasov-Poisson system Introuction to the Vlasov-Poisson system Simone Calogero 1 The Vlasov equation Consier a particle with mass m > 0. Let x(t) R 3 enote the position of the particle at time t R an v(t) = ẋ(t) = x(t)/t its

More information

LATTICE-BASED D-OPTIMUM DESIGN FOR FOURIER REGRESSION

LATTICE-BASED D-OPTIMUM DESIGN FOR FOURIER REGRESSION The Annals of Statistics 1997, Vol. 25, No. 6, 2313 2327 LATTICE-BASED D-OPTIMUM DESIGN FOR FOURIER REGRESSION By Eva Riccomagno, 1 Rainer Schwabe 2 an Henry P. Wynn 1 University of Warwick, Technische

More information

Linear First-Order Equations

Linear First-Order Equations 5 Linear First-Orer Equations Linear first-orer ifferential equations make up another important class of ifferential equations that commonly arise in applications an are relatively easy to solve (in theory)

More information

Affine surface area and convex bodies of elliptic type

Affine surface area and convex bodies of elliptic type Affine surface area and convex bodies of elliptic type Rolf Schneider Abstract If a convex body K in R n is contained in a convex body L of elliptic type (a curvature image), then it is known that the

More information

Table of Common Derivatives By David Abraham

Table of Common Derivatives By David Abraham Prouct an Quotient Rules: Table of Common Derivatives By Davi Abraham [ f ( g( ] = [ f ( ] g( + f ( [ g( ] f ( = g( [ f ( ] g( g( f ( [ g( ] Trigonometric Functions: sin( = cos( cos( = sin( tan( = sec

More information

Math 342 Partial Differential Equations «Viktor Grigoryan

Math 342 Partial Differential Equations «Viktor Grigoryan Math 342 Partial Differential Equations «Viktor Grigoryan 6 Wave equation: solution In this lecture we will solve the wave equation on the entire real line x R. This correspons to a string of infinite

More information

15.1 Upper bound via Sudakov minorization

15.1 Upper bound via Sudakov minorization ECE598: Information-theoretic methos in high-imensional statistics Spring 206 Lecture 5: Suakov, Maurey, an uality of metric entropy Lecturer: Yihong Wu Scribe: Aolin Xu, Mar 7, 206 [E. Mar 24] In this

More information

On the cells in a stationary Poisson hyperplane mosaic

On the cells in a stationary Poisson hyperplane mosaic On the cells in a stationary Poisson hyperplane mosaic Matthias Reitzner and Rolf Schneider Abstract Let X be the mosaic generated by a stationary Poisson hyperplane process X in R d. Under some mild conditions

More information

arxiv: v1 [math.dg] 1 Nov 2015

arxiv: v1 [math.dg] 1 Nov 2015 DARBOUX-WEINSTEIN THEOREM FOR LOCALLY CONFORMALLY SYMPLECTIC MANIFOLDS arxiv:1511.00227v1 [math.dg] 1 Nov 2015 ALEXANDRA OTIMAN AND MIRON STANCIU Abstract. A locally conformally symplectic (LCS) form is

More information

Direct and inverse theorems of approximation theory in L 2 (R d, w l (x)dx)

Direct and inverse theorems of approximation theory in L 2 (R d, w l (x)dx) MATEMATIKA, 2017, Volume 33, Number 2, 177 189 c Penerbit UTM Press. All rights reserve Direct an inverse theorems of approximation theory in L 2 (, w l (x)x) 1 aouan Daher, 2 Salah El Ouaih an 3 Mohame

More information

REAL ANALYSIS I HOMEWORK 5

REAL ANALYSIS I HOMEWORK 5 REAL ANALYSIS I HOMEWORK 5 CİHAN BAHRAN The questions are from Stein an Shakarchi s text, Chapter 3. 1. Suppose ϕ is an integrable function on R with R ϕ(x)x = 1. Let K δ(x) = δ ϕ(x/δ), δ > 0. (a) Prove

More information

GLOBAL SOLUTIONS FOR 2D COUPLED BURGERS-COMPLEX-GINZBURG-LANDAU EQUATIONS

GLOBAL SOLUTIONS FOR 2D COUPLED BURGERS-COMPLEX-GINZBURG-LANDAU EQUATIONS Electronic Journal of Differential Equations, Vol. 015 015), No. 99, pp. 1 14. ISSN: 107-6691. URL: http://eje.math.txstate.eu or http://eje.math.unt.eu ftp eje.math.txstate.eu GLOBAL SOLUTIONS FOR D COUPLED

More information

Section 2.7 Derivatives of powers of functions

Section 2.7 Derivatives of powers of functions Section 2.7 Derivatives of powers of functions (3/19/08) Overview: In this section we iscuss the Chain Rule formula for the erivatives of composite functions that are forme by taking powers of other functions.

More information

Euler equations for multiple integrals

Euler equations for multiple integrals Euler equations for multiple integrals January 22, 2013 Contents 1 Reminer of multivariable calculus 2 1.1 Vector ifferentiation......................... 2 1.2 Matrix ifferentiation........................

More information

Least-Squares Regression on Sparse Spaces

Least-Squares Regression on Sparse Spaces Least-Squares Regression on Sparse Spaces Yuri Grinberg, Mahi Milani Far, Joelle Pineau School of Computer Science McGill University Montreal, Canaa {ygrinb,mmilan1,jpineau}@cs.mcgill.ca 1 Introuction

More information

1 Math 285 Homework Problem List for S2016

1 Math 285 Homework Problem List for S2016 1 Math 85 Homework Problem List for S016 Note: solutions to Lawler Problems will appear after all of the Lecture Note Solutions. 1.1 Homework 1. Due Friay, April 8, 016 Look at from lecture note exercises:

More information

Lecture 2 Lagrangian formulation of classical mechanics Mechanics

Lecture 2 Lagrangian formulation of classical mechanics Mechanics Lecture Lagrangian formulation of classical mechanics 70.00 Mechanics Principle of stationary action MATH-GA To specify a motion uniquely in classical mechanics, it suffices to give, at some time t 0,

More information

On the number of isolated eigenvalues of a pair of particles in a quantum wire

On the number of isolated eigenvalues of a pair of particles in a quantum wire On the number of isolate eigenvalues of a pair of particles in a quantum wire arxiv:1812.11804v1 [math-ph] 31 Dec 2018 Joachim Kerner 1 Department of Mathematics an Computer Science FernUniversität in

More information

1 Heisenberg Representation

1 Heisenberg Representation 1 Heisenberg Representation What we have been ealing with so far is calle the Schröinger representation. In this representation, operators are constants an all the time epenence is carrie by the states.

More information

Physics 505 Electricity and Magnetism Fall 2003 Prof. G. Raithel. Problem Set 3. 2 (x x ) 2 + (y y ) 2 + (z + z ) 2

Physics 505 Electricity and Magnetism Fall 2003 Prof. G. Raithel. Problem Set 3. 2 (x x ) 2 + (y y ) 2 + (z + z ) 2 Physics 505 Electricity an Magnetism Fall 003 Prof. G. Raithel Problem Set 3 Problem.7 5 Points a): Green s function: Using cartesian coorinates x = (x, y, z), it is G(x, x ) = 1 (x x ) + (y y ) + (z z

More information

arxiv: v4 [cs.ds] 7 Mar 2014

arxiv: v4 [cs.ds] 7 Mar 2014 Analysis of Agglomerative Clustering Marcel R. Ackermann Johannes Blömer Daniel Kuntze Christian Sohler arxiv:101.697v [cs.ds] 7 Mar 01 Abstract The iameter k-clustering problem is the problem of partitioning

More information

On a Generalization of the Busemann Petty Problem

On a Generalization of the Busemann Petty Problem Convex Geometric Analysis MSRI Publications Volume 34, 1998 On a Generalization of the Busemann Petty Problem JEAN BOURGAIN AND GAOYONG ZHANG Abstract. The generalized Busemann Petty problem asks: If K

More information

SYMPLECTIC GEOMETRY: LECTURE 3

SYMPLECTIC GEOMETRY: LECTURE 3 SYMPLECTIC GEOMETRY: LECTURE 3 LIAT KESSLER 1. Local forms Vector fiels an the Lie erivative. A vector fiel on a manifol M is a smooth assignment of a vector tangent to M at each point. We think of M as

More information

Computing Exact Confidence Coefficients of Simultaneous Confidence Intervals for Multinomial Proportions and their Functions

Computing Exact Confidence Coefficients of Simultaneous Confidence Intervals for Multinomial Proportions and their Functions Working Paper 2013:5 Department of Statistics Computing Exact Confience Coefficients of Simultaneous Confience Intervals for Multinomial Proportions an their Functions Shaobo Jin Working Paper 2013:5

More information

7.1 Support Vector Machine

7.1 Support Vector Machine 67577 Intro. to Machine Learning Fall semester, 006/7 Lecture 7: Support Vector Machines an Kernel Functions II Lecturer: Amnon Shashua Scribe: Amnon Shashua 7. Support Vector Machine We return now to

More information

The effect of dissipation on solutions of the complex KdV equation

The effect of dissipation on solutions of the complex KdV equation Mathematics an Computers in Simulation 69 (25) 589 599 The effect of issipation on solutions of the complex KV equation Jiahong Wu a,, Juan-Ming Yuan a,b a Department of Mathematics, Oklahoma State University,

More information

WUCHEN LI AND STANLEY OSHER

WUCHEN LI AND STANLEY OSHER CONSTRAINED DYNAMICAL OPTIMAL TRANSPORT AND ITS LAGRANGIAN FORMULATION WUCHEN LI AND STANLEY OSHER Abstract. We propose ynamical optimal transport (OT) problems constraine in a parameterize probability

More information

Noether s theorem applied to classical electrodynamics

Noether s theorem applied to classical electrodynamics Noether s theorem applie to classical electroynamics Thomas B. Mieling Faculty of Physics, University of ienna Boltzmanngasse 5, 090 ienna, Austria (Date: November 8, 207) The consequences of gauge invariance

More information

Math 1271 Solutions for Fall 2005 Final Exam

Math 1271 Solutions for Fall 2005 Final Exam Math 7 Solutions for Fall 5 Final Eam ) Since the equation + y = e y cannot be rearrange algebraically in orer to write y as an eplicit function of, we must instea ifferentiate this relation implicitly

More information

Lower Bounds for k-distance Approximation

Lower Bounds for k-distance Approximation Lower Bouns for k-distance Approximation Quentin Mérigot March 21, 2013 Abstract Consier a set P of N ranom points on the unit sphere of imension 1, an the symmetrize set S = P ( P). The halving polyheron

More information

19 Eigenvalues, Eigenvectors, Ordinary Differential Equations, and Control

19 Eigenvalues, Eigenvectors, Ordinary Differential Equations, and Control 19 Eigenvalues, Eigenvectors, Orinary Differential Equations, an Control This section introuces eigenvalues an eigenvectors of a matrix, an iscusses the role of the eigenvalues in etermining the behavior

More information

Math Notes on differentials, the Chain Rule, gradients, directional derivative, and normal vectors

Math Notes on differentials, the Chain Rule, gradients, directional derivative, and normal vectors Math 18.02 Notes on ifferentials, the Chain Rule, graients, irectional erivative, an normal vectors Tangent plane an linear approximation We efine the partial erivatives of f( xy, ) as follows: f f( x+

More information

Robust Forward Algorithms via PAC-Bayes and Laplace Distributions. ω Q. Pr (y(ω x) < 0) = Pr A k

Robust Forward Algorithms via PAC-Bayes and Laplace Distributions. ω Q. Pr (y(ω x) < 0) = Pr A k A Proof of Lemma 2 B Proof of Lemma 3 Proof: Since the support of LL istributions is R, two such istributions are equivalent absolutely continuous with respect to each other an the ivergence is well-efine

More information

Physics 5153 Classical Mechanics. The Virial Theorem and The Poisson Bracket-1

Physics 5153 Classical Mechanics. The Virial Theorem and The Poisson Bracket-1 Physics 5153 Classical Mechanics The Virial Theorem an The Poisson Bracket 1 Introuction In this lecture we will consier two applications of the Hamiltonian. The first, the Virial Theorem, applies to systems

More information

The Sokhotski-Plemelj Formula

The Sokhotski-Plemelj Formula hysics 25 Winter 208 The Sokhotski-lemelj Formula. The Sokhotski-lemelj formula The Sokhotski-lemelj formula is a relation between the following generalize functions (also calle istributions), ±iǫ = iπ(),

More information

Integration Review. May 11, 2013

Integration Review. May 11, 2013 Integration Review May 11, 2013 Goals: Review the funamental theorem of calculus. Review u-substitution. Review integration by parts. Do lots of integration eamples. 1 Funamental Theorem of Calculus In

More information

MINKOWSKI PROBLEM FOR POLYTOPES. α i u i = 0.

MINKOWSKI PROBLEM FOR POLYTOPES. α i u i = 0. ON THE L p MINKOWSKI PROBLEM FOR POLYTOPES DANIEL HUG, ERWIN LUTWAK, DEANE YANG, AND GAOYONG ZHANG Abstract. Two new approaches are presented to establish the existence of polytopal solutions to the discrete-data

More information

Proof of SPNs as Mixture of Trees

Proof of SPNs as Mixture of Trees A Proof of SPNs as Mixture of Trees Theorem 1. If T is an inuce SPN from a complete an ecomposable SPN S, then T is a tree that is complete an ecomposable. Proof. Argue by contraiction that T is not a

More information

A Sketch of Menshikov s Theorem

A Sketch of Menshikov s Theorem A Sketch of Menshikov s Theorem Thomas Bao March 14, 2010 Abstract Let Λ be an infinite, locally finite oriente multi-graph with C Λ finite an strongly connecte, an let p

More information

4.2 First Differentiation Rules; Leibniz Notation

4.2 First Differentiation Rules; Leibniz Notation .. FIRST DIFFERENTIATION RULES; LEIBNIZ NOTATION 307. First Differentiation Rules; Leibniz Notation In this section we erive rules which let us quickly compute the erivative function f (x) for any polynomial

More information

FURTHER BOUNDS FOR THE ESTIMATION ERROR VARIANCE OF A CONTINUOUS STREAM WITH STATIONARY VARIOGRAM

FURTHER BOUNDS FOR THE ESTIMATION ERROR VARIANCE OF A CONTINUOUS STREAM WITH STATIONARY VARIOGRAM FURTHER BOUNDS FOR THE ESTIMATION ERROR VARIANCE OF A CONTINUOUS STREAM WITH STATIONARY VARIOGRAM N. S. BARNETT, S. S. DRAGOMIR, AND I. S. GOMM Abstract. In this paper we establish an upper boun for the

More information

Calculus of Variations

Calculus of Variations Calculus of Variations Lagrangian formalism is the main tool of theoretical classical mechanics. Calculus of Variations is a part of Mathematics which Lagrangian formalism is base on. In this section,

More information

Problem set 2: Solutions Math 207B, Winter 2016

Problem set 2: Solutions Math 207B, Winter 2016 Problem set : Solutions Math 07B, Winter 016 1. A particle of mass m with position x(t) at time t has potential energy V ( x) an kinetic energy T = 1 m x t. The action of the particle over times t t 1

More information

LOCAL AND GLOBAL MINIMALITY RESULTS FOR A NONLOCAL ISOPERIMETRIC PROBLEM ON R N

LOCAL AND GLOBAL MINIMALITY RESULTS FOR A NONLOCAL ISOPERIMETRIC PROBLEM ON R N LOCAL AND GLOBAL MINIMALITY RSULTS FOR A NONLOCAL ISOPRIMTRIC PROBLM ON R N M. BONACINI AND R. CRISTOFRI Abstract. We consier a nonlocal isoperimetric problem efine in the whole space R N, whose nonlocal

More information

Gaussian Measure of Sections of convex bodies

Gaussian Measure of Sections of convex bodies Gaussian Measure of Sections of convex bodies A. Zvavitch Department of Mathematics, University of Missouri, Columbia, MO 652, USA Abstract In this paper we study properties of sections of convex bodies

More information

Schrödinger s equation.

Schrödinger s equation. Physics 342 Lecture 5 Schröinger s Equation Lecture 5 Physics 342 Quantum Mechanics I Wenesay, February 3r, 2010 Toay we iscuss Schröinger s equation an show that it supports the basic interpretation of

More information

Assignment 1. g i (x 1,..., x n ) dx i = 0. i=1

Assignment 1. g i (x 1,..., x n ) dx i = 0. i=1 Assignment 1 Golstein 1.4 The equations of motion for the rolling isk are special cases of general linear ifferential equations of constraint of the form g i (x 1,..., x n x i = 0. i=1 A constraint conition

More information

arxiv: v1 [math-ph] 5 May 2014

arxiv: v1 [math-ph] 5 May 2014 DIFFERENTIAL-ALGEBRAIC SOLUTIONS OF THE HEAT EQUATION VICTOR M. BUCHSTABER, ELENA YU. NETAY arxiv:1405.0926v1 [math-ph] 5 May 2014 Abstract. In this work we introuce the notion of ifferential-algebraic

More information

Conservation Laws. Chapter Conservation of Energy

Conservation Laws. Chapter Conservation of Energy 20 Chapter 3 Conservation Laws In orer to check the physical consistency of the above set of equations governing Maxwell-Lorentz electroynamics [(2.10) an (2.12) or (1.65) an (1.68)], we examine the action

More information

Multi-View Clustering via Canonical Correlation Analysis

Multi-View Clustering via Canonical Correlation Analysis Technical Report TTI-TR-2008-5 Multi-View Clustering via Canonical Correlation Analysis Kamalika Chauhuri UC San Diego Sham M. Kakae Toyota Technological Institute at Chicago ABSTRACT Clustering ata in

More information