Reconstructions that Combine Cell Average Interpolation with Least Squares Fitting

Size: px
Start display at page:

Download "Reconstructions that Combine Cell Average Interpolation with Least Squares Fitting"

Transcription

1 App Math Inf Sci, No, ) 7 Appied Mathematics & Information Sciences An Internationa Journa Reconstructions that Combine Ce Average Interpoation with Least Squares Fitting Francesc Aràndiga and José Jaime Noguera Departament de Matemàtica Apicada Universitat de Vaència, Spain Received: 9 Jun 5, Revised: 7 Aug 5, Accepted: 8 Aug 5 ubished onine: Jan 6 Abstract: In this paper we present two reconstructions in the ce average framewor of mutiresoution a a Harten The first one combines interpoation and east squares fitting and the second one is based on east squares fitting We study some of their properties as we as its approximation order We aso anayze how different adaptive techniques ENO and SR) can be used within these reconstructions We appy them to noise remova and compare the resuts that we obtain with other existing techniques Keywords: Ce average, Least squares, Interpoation, Noninear mutiresoution, Denoising Introduction A common probem in approximation theory is the reconstruction of a function from a discrete set of data which gives information on the function itsef This information usuay comes either as point-vaues or ce-averages of the function over a finite set of points or ces, respectivey The function is then approximated by an interpoant, that is, another function whose vaues or ce-averages at the given set of points or, respectivey, ces, coincide with those of the origina one The interpoation is a inear procedure of the vaues on the given set of points, but in this case the accuracy of the approximation in the presence of a singuarity is imited by its order, so that any stenci crossing the singuarity wi resut in an unsatisfactory approximation This means that increasing the degree of the poynomia wi produce arger regions of poor accuracy around singuarities The choice of stencis that avoid crossing singuarities, whenever this is possibe, is crucia for the improvement of the accuracy of the approximation This is the ey underying the ENO essentiay non-osciatory) technique, introduced by Harten et a [9] in the context of high resoution shoc capturing HRSC) schemes for conservation aws With the ENO interpoant the region affected by each singuarity is reduced to the interva that contains it, assuming the singuarities are sufficienty we separated It is possibe to improve this resut using the Harten s subce resoution technique SR) [,,7]) If the ocation of the singuarity within the ce or a sufficient good approximation of it) is nown, then the oss of accuracy can be avoided In [] we study the appication of these reconstructions to the case where the data are contaminated with noise On one hand we use interpoatory reconstructions and on the other hand we rey on east squares We do not get good interpoants in a cases This motivates us to introduce new reconstructions using both ideas The first one combines interpoation with approximation in the east squares sense caed in this paper Interp-L S ) and the second reconstruction is based, foowing [], on approximate the data in the east square sense caed in this paper L S ) Here we present these reconstructions in the ce average framewor which is more appropriate in the presence of noise In [] we can see these reconstructions in the point vaue context We aso anayze how can be used combined with different non inear techniques ENO and SR see [7] and [9] resp) The paper is organized as foows: We reca in Section the discrete framewor for mutiresoution introduced by Harten focusing on ce average discretizations We introduce the Interp-LS reconstruction for ce averages in Section There, we aso present how this reconstruction can be combined with non-inear techniques In Section 4 we present the east squares Corresponding author e-mai: arandiga@uves

2 74 F Aràndiga, J J Noguera: Reconstructions that combine ce average reconstruction for ce averages L S C ) Section 5 is devoted to show some numerica experiments where we can see the resuts that we obtain with the techniques presented in this paper Finay, in Section 6 we present the concusions Harten s framewor for mutiresoution anaysis The buiding bocs of a mutiresoution anaysis as described by Harten see [8] and references therein) are a sequence of inear decimation operators {D + }L, a nested sequence of inear vector spaces V +, with D + : V + V which, from a vector of V +, computes a coarser representation of it), and a sequence of prediction operators { + } L,+ : V V + which, from a vector of V obtains an approximation to its finer) representation in V + In most appications, V + is a space of sequences of rea numbers Our description impies that the resoution increases with The decimation and prediction operators have to satisfy the compatibiity condition: D + + I V ) Given v + V +, the prediction error e + Q + v + : I V + + D + )v+ obtains the information in v + V + that cannot be predicted from v D + v+ V by the operator + It is easiy seen that the error vector, e + : Q + v +, beongs to the nu space of D +, denoted by N D + ) []) Let G + : V + N D + ) be the operator which assigns to each vector e + V + the coefficients d + of its representation in terms of a given basis, {µ + j }, of N D + ) V +, and et E + be the canonica injection N D + ) V + Ceary G + E + I N D + ) and E + G + I V + The non-redundant information in the error vector is contained in the set of coefficients {di + }, caed the scae coefficients at eve We note that v + and{v,d + } have the same cardinaity These ingredients aow to construct an aternative representation of a vector v + V + Given v + we evauate: { v D + v+, d + G + I V + + D ) + )v+, and given v and d +, computed by ), the vector v + is recovered by the inverse formua v + + v + E + d + This gives the equivaence between v + and {v,d + } By repeating step ) for v one obtains its corresponding decomposition {v,d } Iterating this process from L to we find that a mutiresoution setting {{V } L,{D + }L } and a sequence of corresponding prediction operators { + } L satisfying ) define an invertibe mutiresoution transform In Harten s framewor, the decimation and prediction operators are buit from a sequence of discretization operators and a sequence of compatibe reconstruction operators Given a function beonging to a certain functiona space F the discretization operator D + : F V + obtains a discrete representation of it at the resoution eve defined by V + Conversey, the reconstruction operator R + : V + F obtains a function in F from discrete vaues in V + The reconstruction operators R + are compatibe with D + provided that D + R + I V + ) is satisfied for each The decimation and prediction operators corresponding to these sequences are obtained as D + D R + and + D + R The compatibiity condition ) is a consequence of ) In Harten s framewor, the discretization process specifies the setting, then the choice of a reconstruction operator defines a mutiresoution transformation whose properties are cosey reated to those of the reconstruction From the point of view of data-compression appications, accuracy of the reconstruction is an important feature Ce average mutiresoution anaysis Let f F L [,]) Considering the set of nested dyadic grids defined by X {x i }N i, N N, x i ih, h N,,,L where N N The ce-average discretization operator D + : F V + is defined in [8] as foows: f i + :D + f) i h + + i x + i The decimation can then be computed by: f i i h fx)dx x i + h + x + fx)dx, i N + 4) fx)dx f + + f + ) 5) Let p i x) the poynomia of degree r nr+ n, nr,n N, such that h i+s x i+s p i x)dx f i+s, s n,,nr, 6) then, the reconstruction is defined as foows: I C r x; f ) p i x), x [x i,x i], i,,n

3 App Math Inf Sci, No, ) / wwwnaturaspubishingcom/journasasp 75 The prediction operator is in this case: + f ) ) D + I C r x; f )) h + + x + + f ) D + I C r h + + x + I C r x; f )dx, ) x; f )) As a consequence of 6) we have I C r x; f )dx + f ) + + f ) ) f i 7) and, from 5), e + + e+ Then, we can define di + e + The agorithms for the direct 8) and inverse 9) transforms in the ce average framewor are: f L M f L { f,d,,d L } Direct) for L,, f i f + + f + ), i N d i + f + + f ), i N end M f L M M f L Inverse) for,,l f + + f ) + di +, i N f + f i f + + f ) di +, i N end 9) 8) Interp-L S reconstruction for ce averages Interp-L S reconstruction for point vaues I L S ) As in [] the Interp-L S reconstruction is defined as foows: Given and the interva Ii [x i,x i ] we construct the poynomia of degree r such that interpoates x i, x i and approximates in the sense of east squares at the n + nr > r nodes x i+ j, j n,,,,,nr If we denote q I L S r x; f ) this poynomia, then the reconstruction is I L S r x; f )q I L S r x; f ), x [x i,x i ] and the prediction is + f ) I L S r x + ; f ) Interp-L S reconstruction for ce averages I L S C ) Let p I L S C r x; f ) r +)a x the poynomia of degree r <n + n r, Π r ) such that and h i+s x i+s h i x i p I L S C r x; f )dx f i ) p I L S C r x)dx f i+s, s n,,nr, s, ) nr,n N, in the east square sense That is, that minimizes ) min x i+s p i x)dx f i+s p i x) Π r h x i+s s n,,nr, s If we define q i x) r x pi L S C y; f )dy r a x + and we assume, without oss of generaity, that i and x i+s s, then ) is equivaent to q ) q ) r Since a f r a, then a f q x) f r r a )x+ a x + x f r + a x + x) Conditions ) are equivaent to q s+) q s) f + r in the east square sense and ) a s+) + s + f s ) r a s+) + s + ) f s f, s n,,nr, s Then we have to obtain â R r such that minimizes where: A I L S C r r A I L S C min â R r â ) ˆf n+nr f, n+ ) n) n+ ) r n) r ) ) r r r nr+ ) nr) nr+ ) r nr) r, )

4 76 F Aràndiga, J J Noguera: Reconstructions that combine ce average â [a,,a r ] T, ˆf [ f n,, f, f,, f nr ]T and n,,) }{{} T n Then, using the norma equations, ) T â A I L S C r A I L S C r T ˆf n+nr f ), A I L S C r a f r T â Once we have {a } r define the reconstruction as r and pi L S C x; f ) we I L S C r x; f ) p I L S C r x; f ), x [x i,x i ], 4) and the prediction + f ) h + nr n + f ) h + f i + x + γ r ) f + x + + I L S C r x; f )dx i+, I L S C r x; f )dx f ) Tabe : Fiters, γ r ), of the approximations to f + obtained with the I L S C r reconstruction + f ) nr n γr ) f i+ 4 4 γ, ) γ, ) γ 4, ) γ, ) γ, ) γ,4 ) γ, ) γ, ) γ 4,4 ) In Tabe, we show the fiters, γ r ), for different degrees of the poynomia, r, and different sizes of stenci As in [] and [] we can prove the foowing resuts: roposition The matrix A I L S C r, r>, n+nr>r, defined in ), has fu ran roofwe wi prove that A I L S C r has fu ran, ie rana I L S C r ) r, by induction on the degree of the poynomia p I L S C r When r we have a singe vector, so the ran is Let r N, r >, and assume that rana I L S C r ) r We want to see whether rana I L S C r ) r, n+ n+a,nr+b a+n+b>r, a, b, a N, b N Since rana I L S C r ) r, there is a submatrix M r of A I L S C r such that M r Without oss of generaity, we suppose that n, a,botherwise we coud use the ast row and appy the same reasoning) Then A I L S C r n+,nr n) n ) n) r+ n ) r+ A I L S C r nr+ ) r+ nr r+ Given that n and r : n) r+ n ) r+ ) r+ r+ r+ )n r+ ) 5) Then, n) r+ n ) r can be used as the pivot eement to nuify the eements of its coumn If we denote N r as the resut of appying to M r the inear transformations that enabe nuify such eements, we can define the r r matrix: a a r n)r+ n ) r+ A r N r, where its first row eements a, j A r are a, j a, j A I L S C r n+,nr, such that a i, j M r Therefore, by 5) and the induction hypothesis, A r ) r+ n) r+ n ) r ) N r ) r+ n) r+ n ) r ) M r, and then, rana I L S C r n+,nr )r Finay, by the principe of induction we have the resut roposition The Interp-L S reconstruction for ce averages has a unique soution roofsince the matrix A I L S C r, defined in ), has fu ran by roposition, foowing section, the Interp-L S reconstruction for ce averages has a unique soution roposition If p I L S C r x; f ) is the poynomia I L S C of degree r, the Interp-L S reconstruction for ce averages recovers exacty poynomias of degree s, s r

5 App Math Inf Sci, No, ) / wwwnaturaspubishingcom/journasasp 77 rooflet px) s b x be a poynomia of degree s, with s r We define{ f j } m j, m>r, as: and j+ j px) f j, j,,m, 6) qx) px) s a x +, 7) where a b +,,,s Thans to 6) we have that A I L S C s,m â ˆf m f, where â[a,,a s ] T, ˆf [ f,, f m ]T and A I L S C s,m is defined in ) Then A I L S C s,m â ˆf m f ) 8) On the other hand, we now that the Interp-L S probem for ce averages has a unique soution by roposition In particuar, whether p I L S C r r i,,m + )a x is the soution of the Interp-L S probem defined by ) and ), with { f j } m j, m > r, defined by 6), and since qx) is the soution of this probem by 8), we have that p I L S C qx), and therefore the Interp-L S reconstruction recovers exacty qx) Coroary The order of the Interp-L S reconstruction for ce averages is r Noninear Reconstructions for ce averages When using the Interp-L S reconstruction for ce averages we have that if r increases, r mnr+n+, the interpoation process has higher order accuracy, ie the detais di + wi be smaer when f is smooth on [x i n,x i+nr ] On the other hand, the interva [x i n,x i+nr ] gets arger with nr,n so that a singuarity in x i,x i ) wi affect more detai coefficients Non-inear essentiay non-osciatory ENO) interpoation techniques, which were firsty introduced in [9], circumvent this drawbac ENO Interp-L S reconstruction for ce averages EI L S C ) The idea of ENO interpoation technique is to repace in 4) the poynomia p I L S C r x; f ) by p I L S C r i,n i,m n i x; f ) seected among the m+ poynomias {p I L S C r i,s,m s x; f )} m s, where m nr+n+, in order to avoid the infuence of the singuarity In [,8,9] the seection process is made by picing the east osciatory poynomia using numerica information on the divided differences of f at the points x That is, using the stenci that minimizes f[x i,,x i+m ], m Here, inspired by [], we use a seection which gives us better resuts in the presence of noise First, we define E i,s,m) m i s+ h x i s+ and now ) p I L S C r i,s,m s x, f )dx f i s+ 9) E i,n i,m)min s,,m {E i,s,m)} ) Once this seection is made, we thus define EI L S C r x; f ) p I L S C r i,n i,m n i x; f ), x [x i,x i] and + f ) h + + f ) f i + x + EI L S C r x; f )dx, + f ) SR Interp-L S reconstruction for ce averages SI L S C ) The ENO interpoatory technique sti produces arge detais di + when a singuarity is contained in the interva [x i,x i ] In order to reduce further the interpoation error, subce resoution methods SR) were introduced in [7] Now et s see how we can appy the SR technique [,7]) in the Interp-L S reconstruction for ce averages: Assume that fx) has a jump in [x i,x i ] Then the primitive function of fx), Fx) fy)dy C[,]), has a corner a discontinuity in the derivative) there Note that the sets { f i }N i and F {Fi }N i are equivaents due to the reations F i Fx i ) i fy)dyh i j f j and f j F j F j h Then, as in [] or [], we can use the function ) to detect where the singuarity is The steps to appy the SR technique in the Interp-L S reconstruction for ce averages are: - Taing stencis with m n + nr+ nodes, we cacuate the stenci ENO as in the previous section - If n i m and n i+ the stencis for the ces Ii and I i+ are disjoint We abe the ce as suspect of containing a discontinuity

6 78 F Aràndiga, J J Noguera: Reconstructions that combine ce average - For each suspicious ce we define the function G I i L S C x)q I i+,,m x;f L S r ) q I i,m,x;f L S r ) ) If G I i L S C x i ) GI i L S C x i ) < we abe the ce Ii as singuar 4- If G I i L S C x i ) GI i L S C x + ) >, the node x + ies to the eft of the discontinuity Then + f ) is obtained as foows + f ) qi L S r i,m, x+ ;F ) qi,m, I L S rx+ ;F ) h + ) + γn+nr+, r n nr f i+, ) f ) f + i + f ), ) where some of these coefficients are in Tabe use ) because the reconstruction is consistent due to the interpoatory conditions: + f ) + + f + f h + The same coud be said for 5) f ) F i + Fi f i h Thans to the interpoatory conditions, G I i L S C can be expressed in terms of { f i } see [] for detais) For instance, the fiters of ) for n + nr 4 and r are: G I i L S C x i ) h f i 6 f i f i f i f i f i+5 ), G I i L S C x + ) h 6 89 f i f i f i f i 48 5 f i + f i 5 f 48 i f i f i f i f i+5 ), G I i L S C x i ) h f i f i f i f i 6 f i + f i ) Tabe : Fiters, γ r ), of the approximations to f +, obtained with the SI L S C r n+nr+, reconstruction + f ) n nr γr n+nr+, ) f i ) γ, 9 ) γ5, ) 4 γ5, ) 4 γ7, ) 56 γ5, ) 5 γ7, In the other case G I i L S C ) < is ocated to the right of the discontinuity and ), x + + f ) qi L S r + i+,,m x+ n+nr+ β,n+nr+ r x i ) GI L S C i x + ;F ) qi+,,m I L S rx+ ;F ), h + ) f + i+ 4) f ) f + i + f ), 5) where β r,n+nr+ ) γ r n+nr+, ),,,n+ nr+ + f ) ;F ) q I i,m, L S rx+ ;F )) We can Remar) can be cacuated as: h + q I L S r i+,,m x+ 4 Least squares reconstruction for ce averages L S C ) Let p L S C r x; f ) r + )a x the poynomia of the degree r such that h i+s x i+s p L S C r x; f )dx f i+s, s n,,nr, 6) in the east square sense If we define q i x) r x pl S C y; f )dy r a x + and we assume, without oss of generaity, that i and x i+s s, then 6) is equivaent to q s+) q s) and r r a s+) + s +) f s 7) a s+) + s +) f s, s n,,nr Then we have to obtain â R r that minimizes A r L S C min â R r â ˆf, where: A L S C r n+ ) n) n+ ) r n) r nr+ ) nr) nr+ ) r nr) r 8)

7 App Math Inf Sci, No, ) / wwwnaturaspubishingcom/journasasp 79 with â[a,,a r ] T and ˆf [ f n,, f nr] T Then, using the norma equations, ) T â A L S C r A L S C r T ˆf A L S C r Once we have {a } r the reconstruction r and pl S C x; f ) we define L S C r x x; f i+s ) p L S C r x x; f )dx, x [x i,x i ], i+s and the prediction + f ) h + nr n + f ) h + + x + L S C r x; f )dx γ r ) f i+, 9) + x + L S C r x; f )dx nr β r ) f i+ ) n In Tabe and 4 we can see some of these coefficients Tabe : Fiters, γ r ), of the approximations to f + obtained with the L S C r reconstruction, + f ) nr n γr ) f i+ 4 4 γ, ) γ, ) γ, ) γ 4, ) γ, ) γ, ) γ, ) γ,4 ) γ 4, ) γ 4 4,4 ) Tabe 4: Fiters, β r ), of the approximations to f + f ) obtained with the L S C r reconstruction, + nr n β r ) f i+ 4 4 β, ) β, ) β, ) β 4, ) β, ) β, ) β, ) β,4 ) β 4, ) β 4 4,4 ) Observe that, since i h p x i x)dx f i, this i reconstruction do not satisfy ) and ) Then 7) is not f ) d + i true and f i f + + Then, if we define, in 9), f + the inverse transform is M f L M M f L + for,,l f + + f ) + di +, i N f + + f ) di +, i N end 7 64 f ) d + i, ) and the reconstruction is not consistent We ca it L S C NC To ensure the consistency we can use f + f i f + In this case the inverse transform is: M f L M M f L for,,l f + + f ) + di +, i N f + f i f +, i N ) end and we ca it L S C Is we nown by the east squares probem, that the matrix A L S C r, defined in 8), has fu ran Therefore, simiary to roposition we can prove the resuts: roposition 4If p L S C r x; f) is the poynomia L S C of degree r, the L S reconstruction for ce averages recovers exacty poynomias of degree s, s r Coroary The order of the L S reconstruction for ce averages is r 4 ENO and SR L S reconstruction for ce averages EL S C and SL S C ) Whie ENO can be appied as in section obtaining the EL S C reconstruction), we have a probem appying SR in this case The fact that we do not have the consistency impies that we do not now the point vaues of the primitive function as in Section ) and then we can not ocate where the discontinuity is A way to appy the SR in this context is using the function G I i L S C x) ) to detect the singuarity as in

8 8 F Aràndiga, J J Noguera: Reconstructions that combine ce average section ) and then, if the jump is in [x + evauate 9) + f ) + h + + x + n nr f ) f i + and in the other case ) + f ) h + + n+nr+ + f ) f i +,x+ ], we L S C r n+nr+, x; f )dx γ r n+nr+, ) f i+, ) + f ), 4) x + L S C r,n+nr+, x; f )dx 5) β r,n+nr+ ) f + i+, + f ) 6) In Tabe 5 we show some of the coefficients β and γ r n+nr+, ) β r,n+nr+ ),,,n+ nr+ We ca this reconstruction SL S C Notice that it is consistent because we force it by 4) and 6) We aso can set a non consistent reconstruction as in ), denoted by SL S C NC and defined as SL S C but introducing the foowing changes: Repacing 4) by: + f ) S r pli,,n+nr+ x+ ;F ) p L i,n+nr+, S r x+ ;F ))/h + ), 7) and repacing 6) by: + f ) S r pli,,n+nr+ x+ ;F ) p L i,n+nr+, S r x+ ;F ))/h + 8) ), where p L S r x;f ) denotes the L S reconstruction for point vaues Its definition is the same as in the I L S reconstruction section ) but here the poynomia approximates in the sense of east squares at the m > r + nodes x i+ j, j n,,nr see [] for detais) 4 Combining SI L S C and SL S C Consistent reconstructions recover propery the jumps but with non consistent reconstructions the exact ocation of them is ost On the other hand, SL S C NC is expected to produce smoother resuts than SL S C, because 4) and 6) wi produce osciations in the soutions In order to tae advantage of both schemes we create a combined agorithm between SI L S C and SL S C NC denoted as SI L S C L S C The idea is through the agorithm of SI L S C introduce a second function that cacuates the SL S C NC reconstruction Tabe 5: Fiters, β,n+nr+ r ), of the approximations to f + obtained with the SL S C r,n+nr+ reconstruction + f ) n+nr+ β,n+nr+ r ) f i ) β, ) 4 β, ) 4 β, ) β, The SR decision wi be done by the SI L S C scheme, avoiding osing the ocation of jumps but enabing a east squares approximation both in f and f To sum up, initiay we set { fi L}N L i {gl i }N L i, and introduce the SL S C reconstruction in the SI L S C SR oop: The function G I i L S C x) ) is used to detect the singuarity as in section ) and then, if the jump is in[x +,x+ ], we cacuate: + f ) as in ), + f ) f i + + f ), + ḡ ) as in ) substituting 9) f by ḡ, + ḡ ) as in 7) repacing f by ḡ In the other case the jump is in [x +,x+ ] and: + f ) as in 4), + f ) f i + + f ), + ḡ ) as in 5) substituting 4) f by ḡ, + ḡ ) as in 8) repacing f by ḡ 5 Numerica experiments The purpose of this section is to show how the reconstructions presented in this paper can be used to remove noise We consider the function gx) 4x 5 sin 4x π 5 ) ) if x< 9 π, 4) sinπ 4x 5 )+ π π if 9 x We discretize gx) using 4) with N L 5+L obtaining ḡ L {ḡ L i }N L i Then we obtain f L ḡ L + n, where n is some white Gaussian noise We consider the Signa-to-Noise Ratio, expressed in db, to measure the noise of the signa: SNRg, f) : og N L i ḡl i ) N L i ḡl i f i L) Cacuated with the function awgn of MATLAB R

9 App Math Inf Sci, No, ) / wwwnaturaspubishingcom/journasasp 8 We fix SNR5dB and appy L eves of the the direct transform 8), with different reconstructions, to { f i L}N L i, obtaining { f i }N L/ L i Subsequenty, we eiminate a the detais {d } L and we appy L eves of an inverse transform 9), ) or )) obtaining { ˆf L } N L i Then, we evauate the Root Mean Squared Error, RMSE ˆf L,ḡ L ) N L ˆf i L ḡ L i N ) L i In Figure -a) we can see the resuts that we obtain with the SI L S C reconstruction Section ), using L 5, m 7 and r The noise and the Gibbs phenomenon have been competey removed Now we consider the SL S C reconstruction Section 4), with L 5, m 7 and r, where the position of the singuarity is detected with the function G I i L S C x), ) The imposition of consistency inverse transform )) causes that noise remova is reduced Figure -b)) In Figure -c) we use the SL S C NC reconstruction Section 4) with L5,m7 and r As expected, the ocation of the jump is ost, but the noise reduction improves with respect SL S C In Figure -d), we use the SI L S C L S C reconstruction Section 4) with L 5, m 7 and r Now, the position of the jump is propery detected and the noise reduction maintained Finay, these experimenta resuts are aso compared with other image fitering agorithms VISU [5] and SURE []) from Waveet Shrinage Denoising see p ej [6], [4]), obtaining Figures e) and f) From this experiment we concude that SI L S C is the best choice considering the RMSE, but SI L S C L S C is aso a good choice since it provides a cosey RMSE and visuay smoother soutions In the second experiment we want to see which parameters size of the stenci, m, and degree of the poynomia, r) are more adequate to remove noise We use SI L S C and SI L S C L S C for different noise eves from to db) and we evauate the RMSE between the origina data ḡ 5 and the reconstruction ˆf 5 The resuts that we obtain with degree r and different size of stencis m 5,7,9,, are shown in Figure -a) and c) for SI L S C and SI L S C L S C respectivey When we fix m 9 and tae different degrees, r,,4,5, we obtain the Figure -b) and d) for SI L S C and SI L S C L S C respectivey We deduce from Figure that we have better resuts with ow degree poynomias for both reconstructions Regarding the size of the stenci, for SI L S C it is not an important feature, but for SI L S C L S C better resuts are obtained with m 7 After testing with some signas, we propose that r, 7 m 5 is a genera good choice a) SI L S C ; RMSE75 5 b) SL S C ; RMSE9 5 c) SL S C NC; RMSE768 5 d)si L S C L S C ; RMSE64 5 e) VISU; RMSE85 5 f) SURE; RMSE57 Fig : We appy 5 eves of decimation and 5 eves of reconstruction without detais We use r and m7 from a) to d) Dash-dot ine represent the origina data{ḡ 5 }, the dots are the data with noise { f 5 } and the soid ine the reconstruction

10 8 F Aràndiga, J J Noguera: Reconstructions that combine ce average RMSE RMSE 4 m5 m7 m9 m m SNR 4 a) SI L S C r r r4 r5 In this third experiment we compare the resuts using the reconstructions SI L S C, SI L S C L S C both with m and r ) and SURE We decimate L eves and we eiminate a the detais {d } L Afterwards, we appy L eves of an inverse transform In Figure -a) we show the resuts that we obtain when L5 and in Figure -b) when L We observe that we obtain better resuts with the SI L S C L S C reconstruction, speciay when L 4 SILSC SILSC LSC SURE RMSE RMSE SNR 4 b) SI L S C m5 m7 m9 m m SNR a) SILSC SILSC LSC SURE RMSE RMSE SNR c) SI L S C L S C r r r4 r SNR b) Fig : Comparison between the methods SI L S C, SI L S C L S C eiminating {di }L ) and SURE a) L5, b) L SNR d) SI L S C L S C Fig : For a fixed reconstruction indicated in each picture), and different magnitude of noise we appy 5 eves of decimation and 5 eves of reconstruction without detais RMSE that we obtain is shown, with degree r and different size of stencis m 5,7,9,, for a) and c); and m 9 and different degrees, r,,4,5 for b) and d) Finay, we show in Figure 4 the imit functions of some reconstructions We appy 6 eves of reconstruction to { f i } 5 i with f i, i 7, and f 7 In a) we use I L S C and we obtain a broen ine Despite it means that it woud be inadequate for using in mutiresoution schemes, we have seen in the previous experiments that the resuts are smooth This is because we appy the reconstruction to a sufficienty smoothed signa thans to the ce average decimation In b) we use L S C ), which is consistent The output is extremey broen, due to the imposition of consistency generates peas that are transmitted and increased to the next eve With L S C NC ), Figure 4-c), the imit function is smooth since these peas are not generated

11 App Math Inf Sci, No, ) / wwwnaturaspubishingcom/journasasp 8 Acnowedgement This research was partiay supported by Spanish MTM -74 and MTM a) I L S C b) L S C c) L S C NC Fig 4: Limit functions L6, m7 and r 6 Concusions In this paper we have buit a reconstruction, in the ce average framewor, based on combining interpoation and east squares approximation Interp-L S ) verifying that it meets the requirements for use in mutiresoution schemes a a Harten We have seen how to incude the non-inear techniques ENO and SR in the Interp-L S reconstruction improving the approximation in the vicinity of discontinuities Then, we have presented a reconstruction based on east squares approximations L S ), aso in the ce average framewor In this case we do not have the consistency which is necessary in the mutiresoution schemes a a Harten We have presented an scheme that circumvents this probem Whie ENO can be appied without probems, a way to appy the SR in this context is using the function G I i L S C x), ), to detect the singuarity as in section ) We have shown how these reconstructions can be used to reduce the noise We have compared the resuts that we obtain with the methods introduced in this paper with other existing techniques We observe that in the presence of jumps the resuts we get with the methods presented in this paper SI L S C and SI L S C L S C ) are more accurate in genera References [] F Aràndiga, A Cohen, R Donat, and N Dyn Interpoation and approximation of piecewise smooth functions SIAM J Numer Ana, 4):4 57, 5 [] F Aràndiga and R Donat Noninear mutiscae decompositions: the approach of A Harten Numer Agorithms, -):75 6, [] F Aràndiga and J J Noguera Signa denoising with Harten s mutiresoution using interpoation and east squares fitting Advances in Differentia Equations and Appications SEMA SIMAI Springer Series, 4:7 45, 4 [4] Taswe C The what, how, and why of waveet shrinage denoising Comput Sci Eng, : 9, [5] D L Donoho Idea spatia adaptation by waveet shrinage Biometria, 8):45 455, 994 [6] D L Donoho Denoising via soft threshoding IEEE Transactions on Information Theory, 4:6 67, 995 [7] A Harten ENO schemes with subce resoution J Comput hys, 8:48 84, 989 [8] A Harten Mutiresoution representation of data: A genera framewor SIAM J Numer Ana, :5 56, 996 [9] A Harten, B Engquist, S Osher, and SR Charavarthy Uniformy high-order accurate essentiay nonosciatory schemes III J Comput hys, 7):, 987 [] I M Johnstone and D L Donoho Adapting to smoothness via waveet shinage Journa of the Statistica Association, 94): 4, 995 [] D Mizrachi Remoiving Noise from Discontinous Data hd thesis, Schoo of Mathematica Sciences Appied Math Department, 99 [] J J Noguera Transformaciones mutiescaa no ineaes hd thesis, Appied Math Department, Univertity of Vaencia, Francesc Arandiga was born in Canas, Spain, in 96 He received the hd degree in mathematics from the University of Vaencia, Vaencia, Spain, in 99 From 994 to he has been associate professor and since fu professor of appied mathematics at the University of Vaencia He has been head and secretary of the Department of Appied Mathematics at the University of Vaencia He has aso been coordinator of the graduate program in Computer Science and Computationa Mathematics His research interest incude image processing, mutiresoution anaysis, and waveets He has pubished more than research artices in reputed internationa journas of mathematica and engineering sciences

12 84 F Aràndiga, J J Noguera: Reconstructions that combine ce average José Jaime Noguera received the hd degree in Mathematics at Universitat de Vaència Spain) in He has pubished research artices in conferences and journas reated to appied mathematics His research interests incude Signa Denoising, Image Compression and Mutiresoution Schemes, among others He currenty wors in the fied of teaching

Lecture Note 3: Stationary Iterative Methods

Lecture Note 3: Stationary Iterative Methods MATH 5330: Computationa Methods of Linear Agebra Lecture Note 3: Stationary Iterative Methods Xianyi Zeng Department of Mathematica Sciences, UTEP Stationary Iterative Methods The Gaussian eimination (or

More information

Theory of Generalized k-difference Operator and Its Application in Number Theory

Theory of Generalized k-difference Operator and Its Application in Number Theory Internationa Journa of Mathematica Anaysis Vo. 9, 2015, no. 19, 955-964 HIKARI Ltd, www.m-hiari.com http://dx.doi.org/10.12988/ijma.2015.5389 Theory of Generaized -Difference Operator and Its Appication

More information

Week 6 Lectures, Math 6451, Tanveer

Week 6 Lectures, Math 6451, Tanveer Fourier Series Week 6 Lectures, Math 645, Tanveer In the context of separation of variabe to find soutions of PDEs, we encountered or and in other cases f(x = f(x = a 0 + f(x = a 0 + b n sin nπx { a n

More information

A Brief Introduction to Markov Chains and Hidden Markov Models

A Brief Introduction to Markov Chains and Hidden Markov Models A Brief Introduction to Markov Chains and Hidden Markov Modes Aen B MacKenzie Notes for December 1, 3, &8, 2015 Discrete-Time Markov Chains You may reca that when we first introduced random processes,

More information

Wavelet shrinkage estimators of Hilbert transform

Wavelet shrinkage estimators of Hilbert transform Journa of Approximation Theory 163 (2011) 652 662 www.esevier.com/ocate/jat Fu ength artice Waveet shrinkage estimators of Hibert transform Di-Rong Chen, Yao Zhao Department of Mathematics, LMIB, Beijing

More information

$, (2.1) n="# #. (2.2)

$, (2.1) n=# #. (2.2) Chapter. Eectrostatic II Notes: Most of the materia presented in this chapter is taken from Jackson, Chap.,, and 4, and Di Bartoo, Chap... Mathematica Considerations.. The Fourier series and the Fourier

More information

4 Separation of Variables

4 Separation of Variables 4 Separation of Variabes In this chapter we describe a cassica technique for constructing forma soutions to inear boundary vaue probems. The soution of three cassica (paraboic, hyperboic and eiptic) PDE

More information

Problem set 6 The Perron Frobenius theorem.

Problem set 6 The Perron Frobenius theorem. Probem set 6 The Perron Frobenius theorem. Math 22a4 Oct 2 204, Due Oct.28 In a future probem set I want to discuss some criteria which aow us to concude that that the ground state of a sef-adjoint operator

More information

ORTHOGONAL MULTI-WAVELETS FROM MATRIX FACTORIZATION

ORTHOGONAL MULTI-WAVELETS FROM MATRIX FACTORIZATION J. Korean Math. Soc. 46 2009, No. 2, pp. 281 294 ORHOGONAL MLI-WAVELES FROM MARIX FACORIZAION Hongying Xiao Abstract. Accuracy of the scaing function is very crucia in waveet theory, or correspondingy,

More information

Course 2BA1, Section 11: Periodic Functions and Fourier Series

Course 2BA1, Section 11: Periodic Functions and Fourier Series Course BA, 8 9 Section : Periodic Functions and Fourier Series David R. Wikins Copyright c David R. Wikins 9 Contents Periodic Functions and Fourier Series 74. Fourier Series of Even and Odd Functions...........

More information

An explicit Jordan Decomposition of Companion matrices

An explicit Jordan Decomposition of Companion matrices An expicit Jordan Decomposition of Companion matrices Fermín S V Bazán Departamento de Matemática CFM UFSC 88040-900 Forianópois SC E-mai: fermin@mtmufscbr S Gratton CERFACS 42 Av Gaspard Coriois 31057

More information

Cryptanalysis of PKP: A New Approach

Cryptanalysis of PKP: A New Approach Cryptanaysis of PKP: A New Approach Éiane Jaumes and Antoine Joux DCSSI 18, rue du Dr. Zamenhoff F-92131 Issy-es-Mx Cedex France eiane.jaumes@wanadoo.fr Antoine.Joux@ens.fr Abstract. Quite recenty, in

More information

Haar Decomposition and Reconstruction Algorithms

Haar Decomposition and Reconstruction Algorithms Jim Lambers MAT 773 Fa Semester 018-19 Lecture 15 and 16 Notes These notes correspond to Sections 4.3 and 4.4 in the text. Haar Decomposition and Reconstruction Agorithms Decomposition Suppose we approximate

More information

Melodic contour estimation with B-spline models using a MDL criterion

Melodic contour estimation with B-spline models using a MDL criterion Meodic contour estimation with B-spine modes using a MDL criterion Damien Loive, Ney Barbot, Oivier Boeffard IRISA / University of Rennes 1 - ENSSAT 6 rue de Kerampont, B.P. 80518, F-305 Lannion Cedex

More information

Partial permutation decoding for MacDonald codes

Partial permutation decoding for MacDonald codes Partia permutation decoding for MacDonad codes J.D. Key Department of Mathematics and Appied Mathematics University of the Western Cape 7535 Bevie, South Africa P. Seneviratne Department of Mathematics

More information

Stat 155 Game theory, Yuval Peres Fall Lectures 4,5,6

Stat 155 Game theory, Yuval Peres Fall Lectures 4,5,6 Stat 155 Game theory, Yuva Peres Fa 2004 Lectures 4,5,6 In the ast ecture, we defined N and P positions for a combinatoria game. We wi now show more formay that each starting position in a combinatoria

More information

C. Fourier Sine Series Overview

C. Fourier Sine Series Overview 12 PHILIP D. LOEWEN C. Fourier Sine Series Overview Let some constant > be given. The symboic form of the FSS Eigenvaue probem combines an ordinary differentia equation (ODE) on the interva (, ) with a

More information

Generalized Bell polynomials and the combinatorics of Poisson central moments

Generalized Bell polynomials and the combinatorics of Poisson central moments Generaized Be poynomias and the combinatorics of Poisson centra moments Nicoas Privaut Division of Mathematica Sciences Schoo of Physica and Mathematica Sciences Nanyang Technoogica University SPMS-MAS-05-43,

More information

CS229 Lecture notes. Andrew Ng

CS229 Lecture notes. Andrew Ng CS229 Lecture notes Andrew Ng Part IX The EM agorithm In the previous set of notes, we taked about the EM agorithm as appied to fitting a mixture of Gaussians. In this set of notes, we give a broader view

More information

Copyright information to be inserted by the Publishers. Unsplitting BGK-type Schemes for the Shallow. Water Equations KUN XU

Copyright information to be inserted by the Publishers. Unsplitting BGK-type Schemes for the Shallow. Water Equations KUN XU Copyright information to be inserted by the Pubishers Unspitting BGK-type Schemes for the Shaow Water Equations KUN XU Mathematics Department, Hong Kong University of Science and Technoogy, Cear Water

More information

Available online at ScienceDirect. Procedia Computer Science 96 (2016 )

Available online at  ScienceDirect. Procedia Computer Science 96 (2016 ) Avaiabe onine at www.sciencedirect.com ScienceDirect Procedia Computer Science 96 (206 92 99 20th Internationa Conference on Knowedge Based and Inteigent Information and Engineering Systems Connected categorica

More information

Combining reaction kinetics to the multi-phase Gibbs energy calculation

Combining reaction kinetics to the multi-phase Gibbs energy calculation 7 th European Symposium on Computer Aided Process Engineering ESCAPE7 V. Pesu and P.S. Agachi (Editors) 2007 Esevier B.V. A rights reserved. Combining reaction inetics to the muti-phase Gibbs energy cacuation

More information

MATH 172: MOTIVATION FOR FOURIER SERIES: SEPARATION OF VARIABLES

MATH 172: MOTIVATION FOR FOURIER SERIES: SEPARATION OF VARIABLES MATH 172: MOTIVATION FOR FOURIER SERIES: SEPARATION OF VARIABLES Separation of variabes is a method to sove certain PDEs which have a warped product structure. First, on R n, a inear PDE of order m is

More information

Numerical Simulation for Optimizing Temperature Gradients during Single Crystal Casting Process

Numerical Simulation for Optimizing Temperature Gradients during Single Crystal Casting Process ISIJ Internationa Vo 54 (2014) No 2 pp 254 258 Numerica Simuation for Optimizing Temperature Gradients during Singe Crysta Casting Process Aeksandr Aeksandrovich INOZEMTSEV 1) Aeksandra Sergeevna DUBROVSKAYA

More information

CHAPTER 2 AN INTRODUCTION TO WAVELET ANALYSIS

CHAPTER 2 AN INTRODUCTION TO WAVELET ANALYSIS CHAPTER 2 AN INTRODUCTION TO WAVELET ANALYSIS [This chapter is based on the ectures of Professor D.V. Pai, Department of Mathematics, Indian Institute of Technoogy Bombay, Powai, Mumbai - 400 076, India.]

More information

THE REACHABILITY CONES OF ESSENTIALLY NONNEGATIVE MATRICES

THE REACHABILITY CONES OF ESSENTIALLY NONNEGATIVE MATRICES THE REACHABILITY CONES OF ESSENTIALLY NONNEGATIVE MATRICES by Michae Neumann Department of Mathematics, University of Connecticut, Storrs, CT 06269 3009 and Ronad J. Stern Department of Mathematics, Concordia

More information

AALBORG UNIVERSITY. The distribution of communication cost for a mobile service scenario. Jesper Møller and Man Lung Yiu. R June 2009

AALBORG UNIVERSITY. The distribution of communication cost for a mobile service scenario. Jesper Møller and Man Lung Yiu. R June 2009 AALBORG UNIVERSITY The distribution of communication cost for a mobie service scenario by Jesper Møer and Man Lung Yiu R-29-11 June 29 Department of Mathematica Sciences Aaborg University Fredrik Bajers

More information

4 1-D Boundary Value Problems Heat Equation

4 1-D Boundary Value Problems Heat Equation 4 -D Boundary Vaue Probems Heat Equation The main purpose of this chapter is to study boundary vaue probems for the heat equation on a finite rod a x b. u t (x, t = ku xx (x, t, a < x < b, t > u(x, = ϕ(x

More information

Research of Data Fusion Method of Multi-Sensor Based on Correlation Coefficient of Confidence Distance

Research of Data Fusion Method of Multi-Sensor Based on Correlation Coefficient of Confidence Distance Send Orders for Reprints to reprints@benthamscience.ae 340 The Open Cybernetics & Systemics Journa, 015, 9, 340-344 Open Access Research of Data Fusion Method of Muti-Sensor Based on Correation Coefficient

More information

XSAT of linear CNF formulas

XSAT of linear CNF formulas XSAT of inear CN formuas Bernd R. Schuh Dr. Bernd Schuh, D-50968 Kön, Germany; bernd.schuh@netcoogne.de eywords: compexity, XSAT, exact inear formua, -reguarity, -uniformity, NPcompeteness Abstract. Open

More information

Improving the Reliability of a Series-Parallel System Using Modified Weibull Distribution

Improving the Reliability of a Series-Parallel System Using Modified Weibull Distribution Internationa Mathematica Forum, Vo. 12, 217, no. 6, 257-269 HIKARI Ltd, www.m-hikari.com https://doi.org/1.12988/imf.217.611155 Improving the Reiabiity of a Series-Parae System Using Modified Weibu Distribution

More information

Algorithms to solve massively under-defined systems of multivariate quadratic equations

Algorithms to solve massively under-defined systems of multivariate quadratic equations Agorithms to sove massivey under-defined systems of mutivariate quadratic equations Yasufumi Hashimoto Abstract It is we known that the probem to sove a set of randomy chosen mutivariate quadratic equations

More information

SUPPLEMENTARY MATERIAL TO INNOVATED SCALABLE EFFICIENT ESTIMATION IN ULTRA-LARGE GAUSSIAN GRAPHICAL MODELS

SUPPLEMENTARY MATERIAL TO INNOVATED SCALABLE EFFICIENT ESTIMATION IN ULTRA-LARGE GAUSSIAN GRAPHICAL MODELS ISEE 1 SUPPLEMENTARY MATERIAL TO INNOVATED SCALABLE EFFICIENT ESTIMATION IN ULTRA-LARGE GAUSSIAN GRAPHICAL MODELS By Yingying Fan and Jinchi Lv University of Southern Caifornia This Suppementary Materia

More information

The EM Algorithm applied to determining new limit points of Mahler measures

The EM Algorithm applied to determining new limit points of Mahler measures Contro and Cybernetics vo. 39 (2010) No. 4 The EM Agorithm appied to determining new imit points of Maher measures by Souad E Otmani, Georges Rhin and Jean-Marc Sac-Épée Université Pau Veraine-Metz, LMAM,

More information

Wavelet Galerkin Solution for Boundary Value Problems

Wavelet Galerkin Solution for Boundary Value Problems Internationa Journa of Engineering Research and Deveopment e-issn: 2278-67X, p-issn: 2278-8X, www.ijerd.com Voume, Issue 5 (May 24), PP.2-3 Waveet Gaerkin Soution for Boundary Vaue Probems D. Pate, M.K.

More information

Research Article Numerical Range of Two Operators in Semi-Inner Product Spaces

Research Article Numerical Range of Two Operators in Semi-Inner Product Spaces Abstract and Appied Anaysis Voume 01, Artice ID 846396, 13 pages doi:10.1155/01/846396 Research Artice Numerica Range of Two Operators in Semi-Inner Product Spaces N. K. Sahu, 1 C. Nahak, 1 and S. Nanda

More information

Math 124B January 31, 2012

Math 124B January 31, 2012 Math 124B January 31, 212 Viktor Grigoryan 7 Inhomogeneous boundary vaue probems Having studied the theory of Fourier series, with which we successfuy soved boundary vaue probems for the homogeneous heat

More information

Math 124B January 17, 2012

Math 124B January 17, 2012 Math 124B January 17, 212 Viktor Grigoryan 3 Fu Fourier series We saw in previous ectures how the Dirichet and Neumann boundary conditions ead to respectivey sine and cosine Fourier series of the initia

More information

NOISE-INDUCED STABILIZATION OF STOCHASTIC DIFFERENTIAL EQUATIONS

NOISE-INDUCED STABILIZATION OF STOCHASTIC DIFFERENTIAL EQUATIONS NOISE-INDUCED STABILIZATION OF STOCHASTIC DIFFERENTIAL EQUATIONS TONY ALLEN, EMILY GEBHARDT, AND ADAM KLUBALL 3 ADVISOR: DR. TIFFANY KOLBA 4 Abstract. The phenomenon of noise-induced stabiization occurs

More information

Coupling of LWR and phase transition models at boundary

Coupling of LWR and phase transition models at boundary Couping of LW and phase transition modes at boundary Mauro Garaveo Dipartimento di Matematica e Appicazioni, Università di Miano Bicocca, via. Cozzi 53, 20125 Miano Itay. Benedetto Piccoi Department of

More information

Absolute Value Preconditioning for Symmetric Indefinite Linear Systems

Absolute Value Preconditioning for Symmetric Indefinite Linear Systems MITSUBISHI ELECTRIC RESEARCH LABORATORIES http://www.mer.com Absoute Vaue Preconditioning for Symmetric Indefinite Linear Systems Vecharynski, E.; Knyazev, A.V. TR2013-016 March 2013 Abstract We introduce

More information

A Simple and Efficient Algorithm of 3-D Single-Source Localization with Uniform Cross Array Bing Xue 1 2 a) * Guangyou Fang 1 2 b and Yicai Ji 1 2 c)

A Simple and Efficient Algorithm of 3-D Single-Source Localization with Uniform Cross Array Bing Xue 1 2 a) * Guangyou Fang 1 2 b and Yicai Ji 1 2 c) A Simpe Efficient Agorithm of 3-D Singe-Source Locaization with Uniform Cross Array Bing Xue a * Guangyou Fang b Yicai Ji c Key Laboratory of Eectromagnetic Radiation Sensing Technoogy, Institute of Eectronics,

More information

Statistical Learning Theory: A Primer

Statistical Learning Theory: A Primer Internationa Journa of Computer Vision 38(), 9 3, 2000 c 2000 uwer Academic Pubishers. Manufactured in The Netherands. Statistica Learning Theory: A Primer THEODOROS EVGENIOU, MASSIMILIANO PONTIL AND TOMASO

More information

A proposed nonparametric mixture density estimation using B-spline functions

A proposed nonparametric mixture density estimation using B-spline functions A proposed nonparametric mixture density estimation using B-spine functions Atizez Hadrich a,b, Mourad Zribi a, Afif Masmoudi b a Laboratoire d Informatique Signa et Image de a Côte d Opae (LISIC-EA 4491),

More information

Gauss Law. 2. Gauss s Law: connects charge and field 3. Applications of Gauss s Law

Gauss Law. 2. Gauss s Law: connects charge and field 3. Applications of Gauss s Law Gauss Law 1. Review on 1) Couomb s Law (charge and force) 2) Eectric Fied (fied and force) 2. Gauss s Law: connects charge and fied 3. Appications of Gauss s Law Couomb s Law and Eectric Fied Couomb s

More information

Mat 1501 lecture notes, penultimate installment

Mat 1501 lecture notes, penultimate installment Mat 1501 ecture notes, penutimate instament 1. bounded variation: functions of a singe variabe optiona) I beieve that we wi not actuay use the materia in this section the point is mainy to motivate the

More information

Paper presented at the Workshop on Space Charge Physics in High Intensity Hadron Rings, sponsored by Brookhaven National Laboratory, May 4-7,1998

Paper presented at the Workshop on Space Charge Physics in High Intensity Hadron Rings, sponsored by Brookhaven National Laboratory, May 4-7,1998 Paper presented at the Workshop on Space Charge Physics in High ntensity Hadron Rings, sponsored by Brookhaven Nationa Laboratory, May 4-7,998 Noninear Sef Consistent High Resoution Beam Hao Agorithm in

More information

Some Measures for Asymmetry of Distributions

Some Measures for Asymmetry of Distributions Some Measures for Asymmetry of Distributions Georgi N. Boshnakov First version: 31 January 2006 Research Report No. 5, 2006, Probabiity and Statistics Group Schoo of Mathematics, The University of Manchester

More information

BP neural network-based sports performance prediction model applied research

BP neural network-based sports performance prediction model applied research Avaiabe onine www.jocpr.com Journa of Chemica and Pharmaceutica Research, 204, 6(7:93-936 Research Artice ISSN : 0975-7384 CODEN(USA : JCPRC5 BP neura networ-based sports performance prediction mode appied

More information

Smoothness equivalence properties of univariate subdivision schemes and their projection analogues

Smoothness equivalence properties of univariate subdivision schemes and their projection analogues Numerische Mathematik manuscript No. (wi be inserted by the editor) Smoothness equivaence properties of univariate subdivision schemes and their projection anaogues Phiipp Grohs TU Graz Institute of Geometry

More information

12.2. Maxima and Minima. Introduction. Prerequisites. Learning Outcomes

12.2. Maxima and Minima. Introduction. Prerequisites. Learning Outcomes Maima and Minima 1. Introduction In this Section we anayse curves in the oca neighbourhood of a stationary point and, from this anaysis, deduce necessary conditions satisfied by oca maima and oca minima.

More information

T.C. Banwell, S. Galli. {bct, Telcordia Technologies, Inc., 445 South Street, Morristown, NJ 07960, USA

T.C. Banwell, S. Galli. {bct, Telcordia Technologies, Inc., 445 South Street, Morristown, NJ 07960, USA ON THE SYMMETRY OF THE POWER INE CHANNE T.C. Banwe, S. Gai {bct, sgai}@research.tecordia.com Tecordia Technoogies, Inc., 445 South Street, Morristown, NJ 07960, USA Abstract The indoor power ine network

More information

DISTRIBUTION OF TEMPERATURE IN A SPATIALLY ONE- DIMENSIONAL OBJECT AS A RESULT OF THE ACTIVE POINT SOURCE

DISTRIBUTION OF TEMPERATURE IN A SPATIALLY ONE- DIMENSIONAL OBJECT AS A RESULT OF THE ACTIVE POINT SOURCE DISTRIBUTION OF TEMPERATURE IN A SPATIALLY ONE- DIMENSIONAL OBJECT AS A RESULT OF THE ACTIVE POINT SOURCE Yury Iyushin and Anton Mokeev Saint-Petersburg Mining University, Vasiievsky Isand, 1 st ine, Saint-Petersburg,

More information

Notes: Most of the material presented in this chapter is taken from Jackson, Chap. 2, 3, and 4, and Di Bartolo, Chap. 2. 2π nx i a. ( ) = G n.

Notes: Most of the material presented in this chapter is taken from Jackson, Chap. 2, 3, and 4, and Di Bartolo, Chap. 2. 2π nx i a. ( ) = G n. Chapter. Eectrostatic II Notes: Most of the materia presented in this chapter is taken from Jackson, Chap.,, and 4, and Di Bartoo, Chap... Mathematica Considerations.. The Fourier series and the Fourier

More information

Stochastic Variational Inference with Gradient Linearization

Stochastic Variational Inference with Gradient Linearization Stochastic Variationa Inference with Gradient Linearization Suppementa Materia Tobias Pötz * Anne S Wannenwetsch Stefan Roth Department of Computer Science, TU Darmstadt Preface In this suppementa materia,

More information

Distributed average consensus: Beyond the realm of linearity

Distributed average consensus: Beyond the realm of linearity Distributed average consensus: Beyond the ream of inearity Usman A. Khan, Soummya Kar, and José M. F. Moura Department of Eectrica and Computer Engineering Carnegie Meon University 5 Forbes Ave, Pittsburgh,

More information

First-Order Corrections to Gutzwiller s Trace Formula for Systems with Discrete Symmetries

First-Order Corrections to Gutzwiller s Trace Formula for Systems with Discrete Symmetries c 26 Noninear Phenomena in Compex Systems First-Order Corrections to Gutzwier s Trace Formua for Systems with Discrete Symmetries Hoger Cartarius, Jörg Main, and Günter Wunner Institut für Theoretische

More information

Fourier Series. 10 (D3.9) Find the Cesàro sum of the series. 11 (D3.9) Let a and b be real numbers. Under what conditions does a series of the form

Fourier Series. 10 (D3.9) Find the Cesàro sum of the series. 11 (D3.9) Let a and b be real numbers. Under what conditions does a series of the form Exercises Fourier Anaysis MMG70, Autumn 007 The exercises are taken from: Version: Monday October, 007 DXY Section XY of H F Davis, Fourier Series and orthogona functions EÖ Some exercises from earier

More information

MARKOV CHAINS AND MARKOV DECISION THEORY. Contents

MARKOV CHAINS AND MARKOV DECISION THEORY. Contents MARKOV CHAINS AND MARKOV DECISION THEORY ARINDRIMA DATTA Abstract. In this paper, we begin with a forma introduction to probabiity and expain the concept of random variabes and stochastic processes. After

More information

BALANCING REGULAR MATRIX PENCILS

BALANCING REGULAR MATRIX PENCILS BALANCING REGULAR MATRIX PENCILS DAMIEN LEMONNIER AND PAUL VAN DOOREN Abstract. In this paper we present a new diagona baancing technique for reguar matrix pencis λb A, which aims at reducing the sensitivity

More information

On Some Basic Properties of Geometric Real Sequences

On Some Basic Properties of Geometric Real Sequences On Some Basic Properties of eometric Rea Sequences Khirod Boruah Research Schoar, Department of Mathematics, Rajiv andhi University Rono His, Doimukh-791112, Arunacha Pradesh, India Abstract Objective

More information

D. Prémel, J.M. Decitre and G. Pichenot. CEA, LIST, F Gif-sur-Yvette, France

D. Prémel, J.M. Decitre and G. Pichenot. CEA, LIST, F Gif-sur-Yvette, France SIMULATION OF EDDY CURRENT INSPECTION INCLUDING MAGNETIC FIELD SENSOR SUCH AS A GIANT MAGNETO-RESISTANCE OVER PLANAR STRATIFIED MEDIA COMPONENTS WITH EMBEDDED FLAWS D. Préme, J.M. Decitre and G. Pichenot

More information

HYDROGEN ATOM SELECTION RULES TRANSITION RATES

HYDROGEN ATOM SELECTION RULES TRANSITION RATES DOING PHYSICS WITH MATLAB QUANTUM PHYSICS Ian Cooper Schoo of Physics, University of Sydney ian.cooper@sydney.edu.au HYDROGEN ATOM SELECTION RULES TRANSITION RATES DOWNLOAD DIRECTORY FOR MATLAB SCRIPTS

More information

A UNIVERSAL METRIC FOR THE CANONICAL BUNDLE OF A HOLOMORPHIC FAMILY OF PROJECTIVE ALGEBRAIC MANIFOLDS

A UNIVERSAL METRIC FOR THE CANONICAL BUNDLE OF A HOLOMORPHIC FAMILY OF PROJECTIVE ALGEBRAIC MANIFOLDS A UNIERSAL METRIC FOR THE CANONICAL BUNDLE OF A HOLOMORPHIC FAMILY OF PROJECTIE ALGEBRAIC MANIFOLDS DROR AROLIN Dedicated to M Saah Baouendi on the occasion of his 60th birthday 1 Introduction In his ceebrated

More information

Explicit overall risk minimization transductive bound

Explicit overall risk minimization transductive bound 1 Expicit overa risk minimization transductive bound Sergio Decherchi, Paoo Gastado, Sandro Ridea, Rodofo Zunino Dept. of Biophysica and Eectronic Engineering (DIBE), Genoa University Via Opera Pia 11a,

More information

Technische Universität Chemnitz

Technische Universität Chemnitz Technische Universität Chemnitz Sonderforschungsbereich 393 Numerische Simuation auf massiv paraeen Rechnern Zhanav T. Some choices of moments of renabe function and appications Preprint SFB393/03-11 Preprint-Reihe

More information

Converting Z-number to Fuzzy Number using. Fuzzy Expected Value

Converting Z-number to Fuzzy Number using. Fuzzy Expected Value ISSN 1746-7659, Engand, UK Journa of Information and Computing Science Vo. 1, No. 4, 017, pp.91-303 Converting Z-number to Fuzzy Number using Fuzzy Expected Vaue Mahdieh Akhbari * Department of Industria

More information

Tracking Control of Multiple Mobile Robots

Tracking Control of Multiple Mobile Robots Proceedings of the 2001 IEEE Internationa Conference on Robotics & Automation Seou, Korea May 21-26, 2001 Tracking Contro of Mutipe Mobie Robots A Case Study of Inter-Robot Coision-Free Probem Jurachart

More information

WAVELET LINEAR ESTIMATION FOR DERIVATIVES OF A DENSITY FROM OBSERVATIONS OF MIXTURES WITH VARYING MIXING PROPORTIONS. B. L. S.

WAVELET LINEAR ESTIMATION FOR DERIVATIVES OF A DENSITY FROM OBSERVATIONS OF MIXTURES WITH VARYING MIXING PROPORTIONS. B. L. S. Indian J. Pure App. Math., 41(1): 275-291, February 2010 c Indian Nationa Science Academy WAVELET LINEAR ESTIMATION FOR DERIVATIVES OF A DENSITY FROM OBSERVATIONS OF MIXTURES WITH VARYING MIXING PROPORTIONS

More information

Reichenbachian Common Cause Systems

Reichenbachian Common Cause Systems Reichenbachian Common Cause Systems G. Hofer-Szabó Department of Phiosophy Technica University of Budapest e-mai: gszabo@hps.ete.hu Mikós Rédei Department of History and Phiosophy of Science Eötvös University,

More information

An Algorithm for Pruning Redundant Modules in Min-Max Modular Network

An Algorithm for Pruning Redundant Modules in Min-Max Modular Network An Agorithm for Pruning Redundant Modues in Min-Max Moduar Network Hui-Cheng Lian and Bao-Liang Lu Department of Computer Science and Engineering, Shanghai Jiao Tong University 1954 Hua Shan Rd., Shanghai

More information

u(x) s.t. px w x 0 Denote the solution to this problem by ˆx(p, x). In order to obtain ˆx we may simply solve the standard problem max x 0

u(x) s.t. px w x 0 Denote the solution to this problem by ˆx(p, x). In order to obtain ˆx we may simply solve the standard problem max x 0 Bocconi University PhD in Economics - Microeconomics I Prof M Messner Probem Set 4 - Soution Probem : If an individua has an endowment instead of a monetary income his weath depends on price eves In particuar,

More information

A Novel Learning Method for Elman Neural Network Using Local Search

A Novel Learning Method for Elman Neural Network Using Local Search Neura Information Processing Letters and Reviews Vo. 11, No. 8, August 2007 LETTER A Nove Learning Method for Eman Neura Networ Using Loca Search Facuty of Engineering, Toyama University, Gofuu 3190 Toyama

More information

SHEARLET-BASED INVERSE HALFTONING

SHEARLET-BASED INVERSE HALFTONING Journa of Theoretica and Appied Information Technoog 3 st March 3. Vo. 49 No.3 5-3 JATIT & LLS. A rights reserved. ISSN: 99-8645 www.atit.org E-ISSN: 87-395 SHEARLET-BASED INVERSE HALFTONING CHENGZHI DENG,

More information

Research Article On the Lower Bound for the Number of Real Roots of a Random Algebraic Equation

Research Article On the Lower Bound for the Number of Real Roots of a Random Algebraic Equation Appied Mathematics and Stochastic Anaysis Voume 007, Artice ID 74191, 8 pages doi:10.1155/007/74191 Research Artice On the Lower Bound for the Number of Rea Roots of a Random Agebraic Equation Takashi

More information

CONJUGATE GRADIENT WITH SUBSPACE OPTIMIZATION

CONJUGATE GRADIENT WITH SUBSPACE OPTIMIZATION CONJUGATE GRADIENT WITH SUBSPACE OPTIMIZATION SAHAR KARIMI AND STEPHEN VAVASIS Abstract. In this paper we present a variant of the conjugate gradient (CG) agorithm in which we invoke a subspace minimization

More information

Seasonal Time Series Data Forecasting by Using Neural Networks Multiscale Autoregressive Model

Seasonal Time Series Data Forecasting by Using Neural Networks Multiscale Autoregressive Model American ourna of Appied Sciences 7 (): 372-378, 2 ISSN 546-9239 2 Science Pubications Seasona Time Series Data Forecasting by Using Neura Networks Mutiscae Autoregressive Mode Suhartono, B.S.S. Uama and

More information

Lower Bounds for the Relative Greedy Algorithm for Approximating Steiner Trees

Lower Bounds for the Relative Greedy Algorithm for Approximating Steiner Trees This paper appeared in: Networks 47:2 (2006), -5 Lower Bounds for the Reative Greed Agorithm for Approimating Steiner Trees Stefan Hougard Stefan Kirchner Humbodt-Universität zu Berin Institut für Informatik

More information

Analysis of Emerson s Multiple Model Interpolation Estimation Algorithms: The MIMO Case

Analysis of Emerson s Multiple Model Interpolation Estimation Algorithms: The MIMO Case Technica Report PC-04-00 Anaysis of Emerson s Mutipe Mode Interpoation Estimation Agorithms: The MIMO Case João P. Hespanha Dae E. Seborg University of Caifornia, Santa Barbara February 0, 004 Anaysis

More information

Integrating Factor Methods as Exponential Integrators

Integrating Factor Methods as Exponential Integrators Integrating Factor Methods as Exponentia Integrators Borisav V. Minchev Department of Mathematica Science, NTNU, 7491 Trondheim, Norway Borko.Minchev@ii.uib.no Abstract. Recenty a ot of effort has been

More information

Finite element method for structural dynamic and stability analyses

Finite element method for structural dynamic and stability analyses Finite eement method for structura dynamic and stabiity anayses Modue-9 Structura stabiity anaysis Lecture-33 Dynamic anaysis of stabiity and anaysis of time varying systems Prof C S Manohar Department

More information

HILBERT? What is HILBERT? Matlab Implementation of Adaptive 2D BEM. Dirk Praetorius. Features of HILBERT

HILBERT? What is HILBERT? Matlab Implementation of Adaptive 2D BEM. Dirk Praetorius. Features of HILBERT Söerhaus-Workshop 2009 October 16, 2009 What is HILBERT? HILBERT Matab Impementation of Adaptive 2D BEM joint work with M. Aurada, M. Ebner, S. Ferraz-Leite, P. Godenits, M. Karkuik, M. Mayr Hibert Is

More information

Approach to Identifying Raindrop Vibration Signal Detected by Optical Fiber

Approach to Identifying Raindrop Vibration Signal Detected by Optical Fiber Sensors & Transducers, o. 6, Issue, December 3, pp. 85-9 Sensors & Transducers 3 by IFSA http://www.sensorsporta.com Approach to Identifying Raindrop ibration Signa Detected by Optica Fiber ongquan QU,

More information

Restricted weak type on maximal linear and multilinear integral maps.

Restricted weak type on maximal linear and multilinear integral maps. Restricted weak type on maxima inear and mutiinear integra maps. Oscar Basco Abstract It is shown that mutiinear operators of the form T (f 1,..., f k )(x) = R K(x, y n 1,..., y k )f 1 (y 1 )...f k (y

More information

Discrete Techniques. Chapter Introduction

Discrete Techniques. Chapter Introduction Chapter 3 Discrete Techniques 3. Introduction In the previous two chapters we introduced Fourier transforms of continuous functions of the periodic and non-periodic (finite energy) type, we as various

More information

Numerical solution of one dimensional contaminant transport equation with variable coefficient (temporal) by using Haar wavelet

Numerical solution of one dimensional contaminant transport equation with variable coefficient (temporal) by using Haar wavelet Goba Journa of Pure and Appied Mathematics. ISSN 973-1768 Voume 1, Number (16), pp. 183-19 Research India Pubications http://www.ripubication.com Numerica soution of one dimensiona contaminant transport

More information

Discrete Techniques. Chapter Introduction

Discrete Techniques. Chapter Introduction Chapter 3 Discrete Techniques 3. Introduction In the previous two chapters we introduced Fourier transforms of continuous functions of the periodic and non-periodic (finite energy) type, as we as various

More information

New Efficiency Results for Makespan Cost Sharing

New Efficiency Results for Makespan Cost Sharing New Efficiency Resuts for Makespan Cost Sharing Yvonne Beischwitz a, Forian Schoppmann a, a University of Paderborn, Department of Computer Science Fürstenaee, 3302 Paderborn, Germany Abstract In the context

More information

Separation of Variables and a Spherical Shell with Surface Charge

Separation of Variables and a Spherical Shell with Surface Charge Separation of Variabes and a Spherica She with Surface Charge In cass we worked out the eectrostatic potentia due to a spherica she of radius R with a surface charge density σθ = σ cos θ. This cacuation

More information

1 Equivalent SDOF Approach. Sri Tudjono 1,*, and Patria Kusumaningrum 2

1 Equivalent SDOF Approach. Sri Tudjono 1,*, and Patria Kusumaningrum 2 MATEC Web of Conferences 159, 01005 (018) IJCAET & ISAMPE 017 https://doi.org/10.1051/matecconf/01815901005 Dynamic Response of RC Cantiever Beam by Equivaent Singe Degree of Freedom Method on Eastic Anaysis

More information

Process Capability Proposal. with Polynomial Profile

Process Capability Proposal. with Polynomial Profile Contemporary Engineering Sciences, Vo. 11, 2018, no. 85, 4227-4236 HIKARI Ltd, www.m-hikari.com https://doi.org/10.12988/ces.2018.88467 Process Capabiity Proposa with Poynomia Profie Roberto José Herrera

More information

Do Schools Matter for High Math Achievement? Evidence from the American Mathematics Competitions Glenn Ellison and Ashley Swanson Online Appendix

Do Schools Matter for High Math Achievement? Evidence from the American Mathematics Competitions Glenn Ellison and Ashley Swanson Online Appendix VOL. NO. DO SCHOOLS MATTER FOR HIGH MATH ACHIEVEMENT? 43 Do Schoos Matter for High Math Achievement? Evidence from the American Mathematics Competitions Genn Eison and Ashey Swanson Onine Appendix Appendix

More information

ON THE REPRESENTATION OF OPERATORS IN BASES OF COMPACTLY SUPPORTED WAVELETS

ON THE REPRESENTATION OF OPERATORS IN BASES OF COMPACTLY SUPPORTED WAVELETS SIAM J. NUMER. ANAL. c 1992 Society for Industria Appied Mathematics Vo. 6, No. 6, pp. 1716-1740, December 1992 011 ON THE REPRESENTATION OF OPERATORS IN BASES OF COMPACTLY SUPPORTED WAVELETS G. BEYLKIN

More information

TWO DENOISING SURE-LET METHODS FOR COMPLEX OVERSAMPLED SUBBAND DECOMPOSITIONS

TWO DENOISING SURE-LET METHODS FOR COMPLEX OVERSAMPLED SUBBAND DECOMPOSITIONS TWO DENOISING SUE-ET METHODS FO COMPEX OVESAMPED SUBBAND DECOMPOSITIONS Jérôme Gauthier 1, aurent Duva 2, and Jean-Christophe Pesquet 1 1 Université Paris-Est Institut Gaspard Monge and UM CNS 8049, 77454

More information

Physics 505 Fall 2007 Homework Assignment #5 Solutions. Textbook problems: Ch. 3: 3.13, 3.17, 3.26, 3.27

Physics 505 Fall 2007 Homework Assignment #5 Solutions. Textbook problems: Ch. 3: 3.13, 3.17, 3.26, 3.27 Physics 55 Fa 7 Homework Assignment #5 Soutions Textook proems: Ch. 3: 3.3, 3.7, 3.6, 3.7 3.3 Sove for the potentia in Proem 3., using the appropriate Green function otained in the text, and verify that

More information

arxiv: v1 [math.co] 17 Dec 2018

arxiv: v1 [math.co] 17 Dec 2018 On the Extrema Maximum Agreement Subtree Probem arxiv:1812.06951v1 [math.o] 17 Dec 2018 Aexey Markin Department of omputer Science, Iowa State University, USA amarkin@iastate.edu Abstract Given two phyogenetic

More information

SINGLE BASEPOINT SUBDIVISION SCHEMES FOR MANIFOLD-VALUED DATA: TIME-SYMMETRY WITHOUT SPACE-SYMMETRY

SINGLE BASEPOINT SUBDIVISION SCHEMES FOR MANIFOLD-VALUED DATA: TIME-SYMMETRY WITHOUT SPACE-SYMMETRY SINGLE BASEPOINT SUBDIVISION SCHEMES FOR MANIFOLD-VALUED DATA: TIME-SYMMETRY WITHOUT SPACE-SYMMETRY TOM DUCHAMP, GANG XIE, AND THOMAS YU Abstract. This paper estabishes smoothness resuts for a cass of

More information

Pattern Frequency Sequences and Internal Zeros

Pattern Frequency Sequences and Internal Zeros Advances in Appied Mathematics 28, 395 420 (2002 doi:10.1006/aama.2001.0789, avaiabe onine at http://www.ideaibrary.com on Pattern Frequency Sequences and Interna Zeros Mikós Bóna Department of Mathematics,

More information

On the Goal Value of a Boolean Function

On the Goal Value of a Boolean Function On the Goa Vaue of a Booean Function Eric Bach Dept. of CS University of Wisconsin 1210 W. Dayton St. Madison, WI 53706 Lisa Heerstein Dept of CSE NYU Schoo of Engineering 2 Metrotech Center, 10th Foor

More information

More Scattering: the Partial Wave Expansion

More Scattering: the Partial Wave Expansion More Scattering: the Partia Wave Expansion Michae Fower /7/8 Pane Waves and Partia Waves We are considering the soution to Schrödinger s equation for scattering of an incoming pane wave in the z-direction

More information