Vertex Unique Labelled Subgraph Mining
|
|
- Cassandra Harrison
- 5 years ago
- Views:
Transcription
1 Verte Unique Lbelled Subgrph Mining Wen Yu, Frns Coenen, Michele Zito, nd Subhieh El Slhi Abstrct With the successful development of efficient lgorithms for Frequent Subgrph Mining (FSM), this pper etends the scope of subgrph mining b proposing Verte Unique lbelled Subgrph Mining (VULSM). VULSM hs focus on the locl properties of grph nd does not require eternl prmeters such s the support threshold used in frequent pttern mining. There re mn pplictions where the mining of VULS is significnt, the ppliction considered in this pper is error prediction with respect to sheet metl forming. More specificll this pper presents formlism for VULSM nd n lgorithm, the Right-most Etension VULS Mining (RE- VULSM) lgorithm, which identifies ll VULS in given grph. The performnce of REVULSM is evluted using rel world sheet metl forming ppliction. The eperimentl results demonstrte tht ll VULS (Verte Unique Lbelled Subgrphs) cn be effectivel identified. 1 Introduction A novel reserch theme in the contet of grph mining [7, 15, 8, 16], Verte Unique lbelled Subgrph Mining (VULSM), is proposed in this pper. Given prticulr sub-grph g in single input grph G; this subgrph will hve specific structure, nd edge nd verte lbelling ssocited with it. If we consider onl the structure nd edge lbelling there m be number of different comptible verte lbellings with respect to G. A Verte Unique Lbelled Subgrph (VULS) is subgrph with specific structure nd edge lbelling tht hs unique verte lbelling ssocited with it. This pper proposes the Deprtment of Computer Science, The Universit of Liverpool Ashton Building, Ashton Street, Liverpool, L69 3BX, UK {uwen,coenen,michele,hsselsl}@liverpool.c.uk
2 Wen Yu, Frns Coenen, Michele Zito, nd Subhieh El Slhi Right-most Etension Verte Unique Lbelled Subgrph Mining Algorithm (REVULSM) to identif ll VULS. REVULSM genertes subgrphs (potentil VULS cndidtes) using Right Most Etension [3], in DFS mnner, s first proposed in the contet of gspn [14]; nd then identifies ll VULS using level-wise pproch (first proposed b Agrwl nd Sriknt in the contet of frequent item set mining [1, 2, 9]). VULSM is pplicble to vrious tpes of grph; however, in this pper we focus on undirected grphs. VULSM hs relevnce with respect to number of domins. The ppliction domin used to illustrte the work described in this pper is error prediction in sheet metl forming. More specificll error prediction in Asmmetric Incrementl Sheet Forming (AISF) [4, 6, 10, 12, 13]. In this scenrio the piece to be mnufctured is represented s grid, ech grid centre point is defined b Eucliden (X-Y-Z) coordinte scheme. The grid cn then be conceptulised s grph (lttice) such tht ech verte represents grid point. Ech verte (ecept t the edges nd corners) cn then be connected to its four neighbours b sequence of edges, which in turn cn be lbelled with slope vlues. An issue with sheet metl forming processes, such s AISF, is tht distortions re introduced s result of the ppliction of the process. These distortions re non-uniform cross the shpe, but tend to be relted to locl geometries. The proposed grph representtion cptures such geometries in terms of sub-grphs, prticulr sub-grphs re ssocited with prticulr locl geometries (nd b etension distortion/error ptterns). Given before nd fter shpes we cn crete trining set b deriving the error ssocited with ech verte in the grid. This trining dt, in turn, cn then be used to trin predictor or clssifier of some sort. There re vrious ws tht such clssifier cn be generted; but one mechnism is to ppl VULSM, s proposed in this pper, to identif sub-grphs tht hve unique error ptterns ssocited with them tht cn then be used for error prediction purposes (some form of mitigting error correction cn then be formulted). A simple emple grid nd corresponding grph re given in Figure 1. The grid (lefthnd side of Figure 1) comprises si grid squres. Ech grid centre is defined b X-Y -Z coordinte tuple. Ech grid centre point is ssocited with verte within the grph (right-hnd side of figure 1). The edges, s noted bove, re lbelled with slope vlues, the difference in the Z coordinte vlues ssocited with the two end vertices. Ech verte will be lbelled with n error vlues (e 1 to e 3 in the figure) describing the epected distortion t tht verte s obtined from trining set (derived from before nd fter grid dt). Identified VULS will describe locl geometries ech with prticulr ssocited error pttern. This knowledge cn then be used to predict errors in unseen grids so tht some form of mitigting error correction cn be pplied. The rest of this pper is orgnised s follows. In Section 2, we define the bsic concepts of VULS together with n illustrtive emple. The REVULSM lgorithm is then described in detil in Section 3. An eperimentl nlsis
3 Verte Unique Lbelled Subgrph Mining 2 1 e 1 e 2 e e 3 e 3 e 2 Fig. 1: Grid representtion (left) with corresponding grph/lttice (right) feturing slope lbels on edges of the pproch is presented in Section 4, nd Section 5 summrises the work nd the min findings, nd presents some conclusions. 2 The problem formultion This section presents forml definition of the concept of VULS. Assume connected lbelled grph G comprised of set of n vertices V, such tht V = {v 1, v 2,..., v n }; nd set of m edges E, such tht E = {e 1, e 2,..., e m }. The vertices re lbelled ccording to set of p verte lbels L V = {l v1, l v2,..., l vp }. The edges re lbelled ccording to set of q edge lbels L E = {l e1, l e2,..., l eq }. A grph G cn thus be conceptulised s comprising k one-edge subgrphs: G = {P 1, P 2,..., P k }, where P i is pir of vertices linked b n edge, thus P i = v, v b (where v, v b V ). The size of grph G ( G ) cn thus be defined in terms of its one edge sub-grphs, we refer to 1-edge subgrphs, 2-edge subgrphs nd so on up to k-edge subgrphs. For undirected grphs, the edge v, v b is equivlent to v b, v (in this pper we ssume undirected subgrphs). We use the nottion P i.v nd P i.v b to indicte the vertices v nd v b ssocited with prticulr verte pir P i, nd the nottion P i.v.lbel nd P i.v b.lbel to indicte the lbels ssocited with P i.v nd P i.v b respectivel. We indicte the sets of lbels which might be ssocited with P i.v nd P i.v b using the nottion L Pi.v nd L Pi.v b (L Pi.v, L Pi.v b L V ). We indicte the edge lbel ssocited with P i using the notion P i.lbel (P i.lbel L E ). We cn use this nottion with respect to n subgrph G sub of G (G sub G). For trining purposes the grphs of interest re required to be lbelled. However, we cn lso conceive of edge onl lbelled grphs nd subgrphs. Given some edge onl lbelled subgrph (G subedgelb ) of some full lbelled grph G (G subedgelb G) comprised of k edges, there m be mn different verte lbelings tht cn be ssocited with such subgrph ccording to the nture of G. We thus define function, getv ertelbels, tht returns the potentil list of lbels S tht cn be ssigned to the vertices in G subedgelb ccording to G:
4 Wen Yu, Frns Coenen, Michele Zito, nd Subhieh El Slhi getv ertelbels(g subedgelb ) S where G subedgelb = {P 1, P 2,..., P k } nd S = [[L P1.v, L P1.v b ], [L P2.v, L P2.v b ],..., [L Pk.v, L Pk.v b ]] (recll tht L Pi.v nd L Pi.v b re the sets of potentil verte lbels for verte v nd v b ssocited with one-edge subgrph P i ). Thus ech element in S comprises two sub-sets of lbels ssocited respectivel with the strt nd end verte for ech edge in G subedgelb ; there is one to one correspondence between ech element (pir of lbel sets) in S with ech element in G subedgelb, hence the re both of the sme size k (recll tht k is the number of edges). We lso ssume tht some cnonicl lbelling is dopted. b b b Fig. 2: Undirected emple lttice b b b Fig. 3: One edge VULS generted from lttice in Figure 2 b Fig. 4: Two edge VULS generted from lttice in Figure 2 b b b Fig. 5: Three edge VULS generted from lttice in Figure 2 According to the bove, the forml definition of the concept of VULS is s follows. Given: (i) k-edge edge lbelled subgrph G subedgelb = {P 1, P 2,..., P k } (G subedgelb G), (ii) list of lbels tht m be ssocited with the vertices in G subedgelb, S = [[L P1.v, L P1.v b ], [L P2.v, L P2.v b ],..., [L Pk.v, L Pk.v b ]], nd (iii) the proviso tht G subedgelb is connected. If [L i, L j ] S, L i = 1, L j = 1 then G subedgelb is k-edge VULS with respect to G. So s to provide for full nd complete comprehension of the concept of VULS n emple lttice is presented in Figure 2. The VULS tht eist in this lttice re itemized in Figures 3 to 5. If we consider one-edge subgrphs first, there re two possibilities: (i) grphs feturing edge, nd (ii) grphs feturing edge. The list of possible vertices S ssocited with the first, obtined using the getv ertelbels function, is [[{}, {}]], while the list ssocited with the second is [[{, b}, {b}]] (this cn be verified b inspection of Figure 2). Considering edge first, [L i, L j ] S, L i = 1 nd L j = 1,
5 Verte Unique Lbelled Subgrph Mining so this is VULS; however, considering edge, [L i, L j ] S, L i 1 nd L j = 1 hence this is not VULS. We now consider the two edge subgrphs b etending the one edge subgrphs. We cn not enumerte ll two edge subgrphs here due to spce limittions but the two edge VULS re shown in Figure 4. Tking the first VULS in Figure 4, {P 1, P 2 }, s n emple, here P 1.v =, P 1.v b =, P 2.v = nd P 2.v b =, furthermore the edge lbels ssocited with P 1 nd P 2 re P 1.lbel = nd P 2.lbel = respectivel. In this cse S = [[, ], [, ]] thus [L i, L j ] S, L i = 1, L j = 1 therefore this is two-edge VULS with respect to G. Algorithm 1 REVULSM 1: Input: 2: G input = Input grph 3: M=M subgrph size 4: Output: 5: R = Set of VULS 6: Globl vribles: 7: G = set of subgrphs (VULS cndidtes) in G input 8: procedure REV ULSM(G input, M) 9: R = 10: G = 11: G = Subgrph Mining(G input ) (Algorithm 2) 12: k=1 13: while (k < M) do 14: for ll G sub G k (where G k is the set of k-edge subgrphs in G) do 15: if IdentifV ULS(G sub, G k ) == true (Algorithm 3) then 16: R = R G sub 17: end if 18: end for 19: k++ 20: end while 21: Return R 22: end procedure 3 The REVULSM lgorithm The proposed REVULSM lgorithm is defined in this section. The lgorithm is founded on the VULS properties presented bove nd mkes use of grph representtion technique borrowed from gspn. The pseudo code for REVULSM is presented in Algorithms 1, 2 nd 3. Algorithm 1 presents the high level control structure, while Algorithm 2 presents the detil for generting ll subgrphs (VULS cndidtes), nd Algorithm 3 the detil for determining whether specific sub-grph is VULS or not.
6 Algorithm 2 Subgrph Mining 1: Input: 2: G input = Input grph 3: Output: 4: G = set of subgrphs in G input 5: Globl vribles: 6: G temp=set of subgrphs generted so fr 7: procedure Subgrph Mining(G input ) 8: G = 9: G temp = 10: G 1 =the set of one-edge subgrphs in G input 11: sort G 1 in DFS leicogrphic order 12: for ech edge e G 1 do 13: G temp = Subgrph(e,1,M) 14: G = G G temp 15: G input = (G input e) (remove e from G input ) 16: end for 17: Return G 18: end procedure Wen Yu, Frns Coenen, Michele Zito, nd Subhieh El Slhi 19: procedure Subgrph(e, size, M) 20: if size > M then 21: return 22: end if 23: generte ll e s potentil etension subgrphs c in G input with one edge growth b right most etension 24: for ech c do 25: if c is miniml DFSCode then 26: G temp = G temp c 27: Subgrph(G temp,size+1,m) 28: end if 29: end for 30: Return G temp 31: end procedure Considering Algorithm 1 first, the lgorithm commences with n input grph G input nd prmeter M tht defines the mimum size of the VULS. If we do not limit the size of the serched-for VULSs the entire input grph m ultimtel be identified s VULS which, in the contet of the sheet metl forming trget ppliction, will not be ver useful. The output is set of VULS R (the set R m include overlps). Note tht ll grphs re encoded using Depth First Serch (DFS) leicogrphicl ordering (s used in gspn [14]). The globl vrible G (line 7 in Algorithm 1) is the set of ll subgrphs in G input. At the strt of the REVULSM procedure, the sets G nd R will be empt. We proceed in depth first mnner to generte ll subgrphs (VULS cndidtes) G b clling lgorithm 2 (line 11). Then we identif VULS from ll subgrphs G strting from one-edge subgrphs (k = 1), then two edge sub-grphs (k = 2), nd so on. We continue in this mnner until k = M
7 Verte Unique Lbelled Subgrph Mining Algorithm 3 IdentifVULS 1: Input: 2: g = single k-edge subgrph (potentil VULS) 3: G k = set of k-edge subgrphs to be compred with g 4: Output: 5: true if g is VULS, flse otherwise 6: procedure IdentifV ULS(g, G k ) 7: isv ULS = true 8: S = the list of potentil verte lbels tht m be ssigned to g 9: for ll [L i, L j ] S do 10: if either L i = 1 or L j = 1 then 11: isv ULS = flse 12: brek 13: end if 14: end for 15: return isv ULS 16: end procedure (line 13-20). On ech itertion lgorithm 3 is clled (line 15) to determine whether G sub is VULS or not with respect to the k-edge subgrphs G k. If it is VULS, it will be dded to the set R. Algorithm 2 comprises two procedures. The first, Subgrph Mining(G input ), is similr to tht found in gspn. We re itertivel finding ll subgrphs, up to size of M. We commence (line 10-11) b sorting ll the one-edge subgrphs, contined in input grph G input, into DFS leicogrphic order nd storing them in G 1. Then (lines 12-16), for ech one edge subgrph e in G 1 we cll the Subgrph procedure (line 13), which finds ll super grphs for ech one edge grph e up to size M in DFS mnner, nd stores the result in G temp ; which is then dded to G (line 14). Finll, we remove e from G input (line 15) to void generting gin n duplicte subgrphs contining e. The Subgrph(e, size, M ) procedure genertes ll the super grphs of the given one edge subgrph e b growing e b dding edges using the right most etension principle. For ech potentil subgrph c, if c is described b miniml DFSCode (line 25) the process is repet (in DFS stle) so s to generte ll the super grphs of e (line 27). The process continues in this recursive mnner until the number of edges in the super grphs to be generted (size) is greter thn M, or no more grphs cn be generted. Algorithm 3 presents the pseudo code for identifing whether given subgrph g is VULS or not with respect to the current set of k-edge sub-grphs G k from which g hs been removed. The lgorithm returns true if g is VULS nd f lse otherwise. The process commences (line 8) b generting the potentil list of verte lbels S tht cn be mtched to g ccording to the content of G k (see previous section for detil). The list S is then processed nd tested. If there eists verte pir whose possible lbelling is not unique
8 Wen Yu, Frns Coenen, Michele Zito, nd Subhieh El Slhi (hs more thn one possible lbelling tht cn be ssocited with it) g is not VULS nd the procedure returns flse, otherwise g is VULS nd the procedure returns true. 4 Eperiments nd Performnce Stud This section describes the performnce stud tht ws conducted to nlse the genertion nd ppliction of the concept of VULS. The reported eperiments were ll pplied to rel ppliction of sheet metl forming, more specificll the ppliction of AISF [11, 5] to the fbriction of flt-topped prmid shpes mnufctured out of sheet steel. This shpe ws chosen s it is frequentl used s benchmrk shpe for conducting eperiments in the contet of AISF (lthough not necessril with respect to error prediction). Nine grphs were generted from this dt using three different grid sizes nd different numbers of edge nd verte lbels; in ddition rnge of vlues were used for the M prmeter. The rest of this sub-section is orgnised s follows. The performnce mesures used with respect to the evlution re itemised in Section 4.1, more detil concerning the dt sets used for the evlution is given in Section 4.2, nd the obtined results re presented nd discussed in Section Eperimentl performnce mesurement Four performnce mesures were used to nlse the effectiveness of the proposed REVULSM : (i) run time (seconds), (ii) number of VULS identified, (iii) discover rte nd (iv) coverge rte. The lst two merit some further eplntion. The discover rte is the rtio of VULS discovered with respect to the totl number of subgrphs of size less thn M (Eqution 1). The coverge rte is the rtio of the number of vertices covered b the detected VULS compred to the totl number of vertices in the input grph (Eqution 2); with respect to the sheet steel forming emple ppliction high coverge rtes re desirble. discover rte (%) = coverge rte (%) = number of V ULS number of subgrphs number of vertices covered b V ULS number of vertices in input grph (1) (2)
9 Verte Unique Lbelled Subgrph Mining 4.2 Dt sets The dt sets used for the evlution consisted of before nd fter coordinte clouds ; the first generted b CAD sstem, the second using n opticl mesuring technique. These were trnsformed into grid representtions, referenced using X-Y-Z coordinte sstem, such tht the before grid could be correlted with the fter grid nd error mesurements obtined. A frgment of the before grid dt, with ssocited error vlues (mm), is presented in Tble 1. The before grid dt ws then trnslted into grph such tht ech grid squre ws represented b verte linked to ech of its neighbouring squres b n edge. Ech verte ws lbelled with n error vlue while the edges were lbelled ccording to the difference in Z of the two end vertices (the slope connecting them). Furthermore, the verte nd edge lbels were discretised so tht the were represented b nominl vlues (otherwise ever edge pir ws likel to be unique). This ws then the input into the REVULSM lgorithm. Tble 1: Formt of rw input dt z error As noted bove, from the rw dt, different sized grid representtions, nd consequentl grph representtions, m be generted. For eperimentl purposes three grid formts were used 6 6, nd We cn lso ssign different numbers of edge lbels to the vertices nd edges, for the evlution reported here vlues of two nd three were used in three different combintions. In totl nine different grph dt sets were generted, numbered AISF1 to AISF9. AISF1 to AISF3 were generted using 6 6 grid, while AISF4 to AISF6 were generted using grid, nd AISF7 to AISF9 were generted using grid. Some sttistics concerning these grph sets re presented in Tble 2.
10 Wen Yu, Frns Coenen, Michele Zito, nd Subhieh El Slhi Tble 2: Summr of AISF grph sets grph # # edge # verte grph # # edge # verte set vertices lbels lbels set vertices lbels lbels AISF AISF AISF AISF AISF AISF AISF AISF AISF Eperimentl results nd nlsis For eperimentl purposes REVULSM ws implemented in the JAVA progrmming lnguge. All eperiments were conducted using 2.7 GHz Intel Core i5 with 4 GB 1333 MHz DDR3 memor, running OS X (12B19). The results obtined re presented in Figures 6 to 17. Figures 6 to 8 give the run time comprisons with respect to the nine grph sets. Figures 9 to 11 give the number of discovered VULS in ech cse. Figures 12 to 14 present comprison of the recorded discover rtes with respect to the nine grph sets considered. Finll Figures 15 to 17 give comprison of the coverge rtes. From Figures 6 to 8 it cn be seen, s might be epected, tht s the vlue of the M prmeter increses the run time lso increses becuse more subgrphs nd hence more VULS re generted. The sme observtion is true with respect to the size of the grph; the more vertices the greter the required runtime. From Figures 9 (6 6 grid), 10 (10 10 grid) nd 11 (21 21 grid) it cn be observed tht s M increses the number of VULS will lso increse, gin this is s might be epected. Compring AISF1, 4 nd 7 with AISF2, 5 nd 8 respectivel, it cn be seen tht s the number of edge lbels increses while the number of verte lbels is kept constnt the number of VULS lso increses (AISF1 nd AISF2 hve the sme number of verte lbels; s do AISF4 nd AISF5, nd AISF7 nd AISF8). This is becuse the likelihood of VULS eisting increses s the input grph becomes more diverse. However, compring AISF2, 5 nd 8 with AISF3, 6 nd 9 respectivel it cn be seen tht s the number of verte lbels increses, while the number of edge lbels is kept constnt, the number of VULS generted will decrese (AISF2 nd AISF3 hve the sme number of edge lbels; s do AISF5 nd AISF6, nd AISF8 nd AISF9); becuse, given high number of verte lbels, the likelihood of VULS eisting decreses. Figure 12 (6 6 grid), 13 (10 10 grid) nd 14 (21 21 grid) show the recorded discover rte vlues. Compring AISF1, 4 nd 7 with AISF2, 5 nd 8 respectivel; when the number of edge lbels increses, while the number of verte lbels is kept constnt, the discover rte increses. This is
11 Verte Unique Lbelled Subgrph Mining Fig. 6: Run time comprison using 3 edge nd 2 verte lbels (AISF1, AISF4 nd AISF7) Fig. 10: Comprison of number of VULS generted (AISF4, AISF5 nd AISF6) Fig. 7: Run time comprison using 2 edge nd 2 verte lbels (AISF2, AISF5 nd AISF8) Fig. 11: Comprison of number of VULS generted (AISF7, AISF8 nd AISF9) Fig. 8: Run time comprison using 2 edge nd 3 verte lbels (AISF3, AISF6 nd AISF9) Fig. 12: Comprison of discover rte (AISF1, AISF2 nd AISF3) Fig. 9: Comprison of number of VULS generted (AISF1, AISF2 nd AISF3) Fig. 13: Comprison of discover rte (AISF4, AISF5 nd AISF6)
12 Wen Yu, Frns Coenen, Michele Zito, nd Subhieh El Slhi Fig. 14: Comprison of discover rte (AISF7, AISF8 nd AISF9) Fig. 16: Comprison of coverge rte (AISF4, AISF5 nd AISF6) Fig. 15: Comprison of coverge rte (AISF1, AISF2 nd AISF3) Fig. 17: Comprison of coverge rte (AISF7, AISF8 nd AISF9) becuse regrdless of the number of edge lbels grph hs (ll other elements being kept constnt) the number of subgrphs contined in the grph will not chnge, while (s indicted b the eperiments reported in Figures 9, 10 nd 11) the number of identified VULS increses s the number of edge lbels increses. Conversel, compring AISF2, 5 nd 8 with AISF3, 6 nd 9 respectivel, when the number of verte lbels increses while the number of edge lbels is kept constnt, the discover rte will decrese becuse (s lred noted) the number of VULS generted decreses s the number of verte lbels increses. It cn lso be noted tht s the M vlue increses, the discover rte does not lws increse, s shown in the cse of AISF4, 5 nd 6. This is becuse s the M vlue increses, the number of VULS goes up s does the number of subgrphs, but the m not both increse t the sme rte. Figures 15, 16 nd 17 show the coverge rte. From the figures it cn be observed tht s M increses the coverge rte lso increses. This is to be epected, however it is interesting to note tht the coverge rte in some cses reches 100% (when M = 6 with respect to AISF1, ASF2 nd AISF4). One hundred percent coverge is desirble in the contet of the sheet metl forming ppliction so tht unique ptterns ssocited with prticulr error distributions (verte lbels) cn be identified for ll geometries. Compring AISF1, 4 nd 7 with AISF2, 5 nd 8 respectivel, the more edge lbels grph hs the more VULS will be generted (see bove); s result more vertices will be covered b VULS nd hence the coverge rte will go up. On the other hnd, compring AISF2, 5 nd 8 with AISF3, 6 nd 9 respectivel,
13 Verte Unique Lbelled Subgrph Mining the more verte lbels grph hs the less VULS will be generted (see bove); s result fewer vertices will be covered b VULS vertices nd hence the coverge rte will go down. 5 Conclusions nd further Stud In this pper we hve proposed the mining of VULS nd presented the RE- VULSM lgorithm. The reported eperimentl results demonstrted tht the VULS ide is sound nd tht REVULSM cn effectivel identif VULS in rel dt. Hving estblished proof on concept there re mn interesting reserch problems relted to VULSM tht cn now be pursued. For instnce, currentl, when the M prmeter is high REVULSM will run out of memor, lthough for the purpose of error prediction in sheet metl forming it cn be rgued tht there is no requirement for lrger VULS, it m be of interest to investigte methods whereb the efficienc of REVULSM cn be improved. Finll, t present, REVULSM finds ll VULS up to predefined size, it is conjectured tht efficienc gins cn be mde if onl miniml VULS re found. 6 Acknowledgements The reserch leding to the results presented in this pper hs received funding from the Europen Union Seventh Frmework Progrmme (FP7/ ) under grnt greement number References 1. Agrwl, R., Sriknt, R.: fst lgorithms for mining ssocition rules. In: Proceedings of the 20th Interntionl Conference on Ver Lrge Dt Bses(VLDB 94), pp (1994) 2. Agrwl, R., Sriknt, R.: Mining sequentil ptterns. In: Proceedings of the Eleventh Interntionl Conference on Dt Engineering(ICDE 95), pp (1995) 3. Asi, T., Abe, K., Kwsoe, S., Skmoto, H., Arikw, S.: Efficient substructure discover from lrge semi-structured dt. In: In Proc.2002 SIAM Int.Conf. Dt Mining, pp (2002) 4. Cfut, G., Mole, N., tok, B.: An enhnced displcement djustment method: Springbck nd thinning compenstion. Mterils nd Design 40, (2012) 5. El-Slhi, S., Coenen, F., Dion, C., Khn, M.S.: Identifiction of correltions between 3d surfces using dt mining techniques: Predicting springbck in sheet metl forming. In: Reserch nd Development in Intelligent Sstems XXIX, pp (2012)
14 Wen Yu, Frns Coenen, Michele Zito, nd Subhieh El Slhi 6. Firt, M., Kftnoglu, B., Eser, O.: Sheet metl forming nlses with n emphsis on the springbck deformtion. Journl of Mterils Processing Technolog 196(1-3), (2008) 7. Hn, J., Cheng, H., Xin, D., Yn, X.: Frequent pttern mining: Current sttus nd future directions. Dt Mining nd Knowledge Discover 15(1), (2007) 8. Hun, J., Wng, W., Prins, J., Yng, J.: SPIN: mining miml frequent subgrphs from grph dtbses. In: Proceedings of the 10th ACM SIGKDD Interntionl Conference on Knowledge Discover nd Dt Mining, pp (2004) 9. Inokuchi, A., Wshio, T., Motod, H.: An priori-bsed lgorithm for mining frequent substructures from grph dt. In: In Principles of Dt Mining nd Knowledge Discover, pp (2000) 10. Jeswiet, J., Micri, F., Hirt, G., Brmle, A., ndj. Allwood, J.D.: Asmmetric single point incrementl forming of sheet metl. CIRP Annls Mnufcturing Technolog 54(2), (2005) 11. Khn, M.S., Coenen, F., Dion, C., El-Slhi, S.: Finding correltions between 3-d surfces: A stud in smmetric incrementl sheet forming. Mchine Lerning nd Dt Mining in Pttern Recognition Lecture Notes in Computer Science 7376, (2012) 12. Liu, W., Ling, Z., Hung, T., Chen, Y., Lin, J.: Process optiml ccontrol of sheet metl forming springbck bsed on evolutionr strteg. In: In Intelligent Control nd Automtion, WCICA th World Congress, pp (June 2008) 13. Nsrollhi, V., Arezoo, B.: Prediction of springbck in sheet metl components with holes on the bending re, using eperiments, finite element nd neurl networks. Mterils nd Design 36, (2012) 14. Yn, X., Hn, J.: gspn: Grph-bsed substructure pttern mining. In: Proceedings of the 2002 Interntionl Conference on Dt Mining, pp (2002) 15. Yn, X., Hn, J.: Close Grph: mining closed frequent grph ptterns. In: Proceedings of the 9th ACM SIGKDD Interntionl Conference on Knowledge Discover nd Dt Mining, pp (2003) 16. Zhu, F., Yn, X., Hn, J., Yu, P.S.: gprune: constrint pushing frmework for grph pttern mining. In: Proceedings of 2007 Pcific-Asi Conference on Knowledge Discover nd Dt Mining (PAKDD 07), pp (2007)
Minimal Vertex Unique Labelled Subgraph Mining
Minimal Vertex Unique Labelled Subgraph Mining Wen Yu, Frans Coenen, Michele Zito, and Subhieh El Salhi Department of Computer Science, The University of Liverpool Ashton Building, Ashton Street, Liverpool,
More information1. SEQUENCES INVOLVING EXPONENTIAL GROWTH (GEOMETRIC SEQUENCES)
Numbers nd Opertions, Algebr, nd Functions 45. SEQUENCES INVOLVING EXPONENTIAL GROWTH (GEOMETRIC SEQUENCES) In sequence of terms involving eponentil growth, which the testing service lso clls geometric
More informationSlides for Data Mining by I. H. Witten and E. Frank
Slides for Dt Mining y I. H. Witten nd E. Frnk Simplicity first Simple lgorithms often work very well! There re mny kinds of simple structure, eg: One ttriute does ll the work All ttriutes contriute eqully
More informationComplete Coverage Path Planning of Mobile Robot Based on Dynamic Programming Algorithm Peng Zhou, Zhong-min Wang, Zhen-nan Li, Yang Li
2nd Interntionl Conference on Electronic & Mechnicl Engineering nd Informtion Technology (EMEIT-212) Complete Coverge Pth Plnning of Mobile Robot Bsed on Dynmic Progrmming Algorithm Peng Zhou, Zhong-min
More informationStained Glass Design. Teaching Goals:
Stined Glss Design Time required 45-90 minutes Teching Gols: 1. Students pply grphic methods to design vrious shpes on the plne.. Students pply geometric trnsformtions of grphs of functions in order to
More informationISG: Itemset based Subgraph Mining
ISG: Itemset bsed Subgrph Mining by Lini Thoms, Stynryn R Vlluri, Kmlkr Krlplem Report No: IIIT/TR/2009/179 Centre for Dt Engineering Interntionl Institute of Informtion Technology Hyderbd - 500 032, INDIA
More informationMidterm 2 Sample solution
Nme: Instructions Midterm 2 Smple solution CMSC 430 Introduction to Compilers Fll 2012 November 28, 2012 This exm contins 9 pges, including this one. Mke sure you hve ll the pges. Write your nme on the
More informationA New Learning Algorithm for the MAXQ Hierarchical Reinforcement Learning Method
A New Lerning Algorithm for the MAXQ Hierrchicl Reinforcement Lerning Method Frzneh Mirzzdeh 1, Bbk Behsz 2, nd Hmid Beigy 1 1 Deprtment of Computer Engineering, Shrif University of Technology, Tehrn,
More informationComputing offsets of freeform curves using quadratic trigonometric splines
Computing offsets of freeform curves using qudrtic trigonometric splines JIULONG GU, JAE-DEUK YUN, YOONG-HO JUNG*, TAE-GYEONG KIM,JEONG-WOON LEE, BONG-JUN KIM School of Mechnicl Engineering Pusn Ntionl
More informationText mining: bag of words representation and beyond it
Text mining: bg of words representtion nd beyond it Jsmink Dobš Fculty of Orgniztion nd Informtics University of Zgreb 1 Outline Definition of text mining Vector spce model or Bg of words representtion
More informationLecture 10 Evolutionary Computation: Evolution strategies and genetic programming
Lecture 10 Evolutionry Computtion: Evolution strtegies nd genetic progrmming Evolution strtegies Genetic progrmming Summry Negnevitsky, Person Eduction, 2011 1 Evolution Strtegies Another pproch to simulting
More informationA TRIANGULAR FINITE ELEMENT FOR PLANE ELASTICITY WITH IN- PLANE ROTATION Dr. Attia Mousa 1 and Eng. Salah M. Tayeh 2
A TRIANGLAR FINITE ELEMENT FOR PLANE ELASTICITY WITH IN- PLANE ROTATION Dr. Atti Mous nd Eng. Slh M. Teh ABSTRACT In the present pper the strin-bsed pproch is pplied to develop new tringulr finite element
More informationTopics in Analytic Geometry
Nme Chpter 10 Topics in Anltic Geometr Section 10.1 Lines Objective: In this lesson ou lerned how to find the inclintion of line, the ngle between two lines, nd the distnce between point nd line. Importnt
More informationFig.1. Let a source of monochromatic light be incident on a slit of finite width a, as shown in Fig. 1.
Answer on Question #5692, Physics, Optics Stte slient fetures of single slit Frunhofer diffrction pttern. The slit is verticl nd illuminted by point source. Also, obtin n expression for intensity distribution
More informationAn Efficient Algorithm for Discovering Frequent Subgraphs. Technical Report
An Efficient Algorithm for Discovering Frequent Sugrphs Technicl Report Deprtment of Computer Science nd Engineering Universit of Minnesot 4-192 EECS Building 200 Union Street SE Minnepolis, MN 55455-0159
More information2014 Haskell January Test Regular Expressions and Finite Automata
0 Hskell Jnury Test Regulr Expressions nd Finite Automt This test comprises four prts nd the mximum mrk is 5. Prts I, II nd III re worth 3 of the 5 mrks vilble. The 0 Hskell Progrmming Prize will be wrded
More informationA Heuristic Approach for Discovering Reference Models by Mining Process Model Variants
A Heuristic Approch for Discovering Reference Models by Mining Process Model Vrints Chen Li 1, Mnfred Reichert 2, nd Andres Wombcher 3 1 Informtion System Group, University of Twente, The Netherlnds lic@cs.utwente.nl
More informationA Transportation Problem Analysed by a New Ranking Method
(IJIRSE) Interntionl Journl of Innovtive Reserch in Science & Engineering ISSN (Online) 7-07 A Trnsporttion Problem Anlysed by New Rnking Method Dr. A. Shy Sudh P. Chinthiy Associte Professor PG Scholr
More informationCHAPTER III IMAGE DEWARPING (CALIBRATION) PROCEDURE
CHAPTER III IMAGE DEWARPING (CALIBRATION) PROCEDURE 3.1 Scheimpflug Configurtion nd Perspective Distortion Scheimpflug criterion were found out to be the best lyout configurtion for Stereoscopic PIV, becuse
More informationUnit #9 : Definite Integral Properties, Fundamental Theorem of Calculus
Unit #9 : Definite Integrl Properties, Fundmentl Theorem of Clculus Gols: Identify properties of definite integrls Define odd nd even functions, nd reltionship to integrl vlues Introduce the Fundmentl
More informationUnit 5 Vocabulary. A function is a special relationship where each input has a single output.
MODULE 3 Terms Definition Picture/Exmple/Nottion 1 Function Nottion Function nottion is n efficient nd effective wy to write functions of ll types. This nottion llows you to identify the input vlue with
More informationA Comparison of the Discretization Approach for CST and Discretization Approach for VDM
Interntionl Journl of Innovtive Reserch in Advnced Engineering (IJIRAE) Volume1 Issue1 (Mrch 2014) A Comprison of the Discretiztion Approch for CST nd Discretiztion Approch for VDM Omr A. A. Shib Fculty
More informationSemistructured Data Management Part 2 - Graph Databases
Semistructured Dt Mngement Prt 2 - Grph Dtbses 2003/4, Krl Aberer, EPFL-SSC, Lbortoire de systèmes d'informtions réprtis Semi-structured Dt - 1 1 Tody's Questions 1. Schems for Semi-structured Dt 2. Grph
More informationMATH 2530: WORKSHEET 7. x 2 y dz dy dx =
MATH 253: WORKSHT 7 () Wrm-up: () Review: polr coordintes, integrls involving polr coordintes, triple Riemnn sums, triple integrls, the pplictions of triple integrls (especilly to volume), nd cylindricl
More informationThe notation y = f(x) gives a way to denote specific values of a function. The value of f at a can be written as f( a ), read f of a.
Chpter Prerequisites for Clculus. Functions nd Grphs Wht ou will lern out... Functions Domins nd Rnges Viewing nd Interpreting Grphs Even Functions nd Odd Functions Smmetr Functions Defined in Pieces Asolute
More informationII. THE ALGORITHM. A. Depth Map Processing
Lerning Plnr Geometric Scene Context Using Stereo Vision Pul G. Bumstrck, Bryn D. Brudevold, nd Pul D. Reynolds {pbumstrck,brynb,pulr2}@stnford.edu CS229 Finl Project Report December 15, 2006 Abstrct A
More informationMatrices and Systems of Equations
Mtrices Mtrices nd Sstems of Equtions A mtri is rectngulr rr of rel numbers. CHAT Pre-Clculus Section 8. m m m............ n n n mn We will use the double subscript nottion for ech element of the mtri.
More informationIntroduction Transformation formulae Polar graphs Standard curves Polar equations Test GRAPHS INU0114/514 (MATHS 1)
POLAR EQUATIONS AND GRAPHS GEOMETRY INU4/54 (MATHS ) Dr Adrin Jnnett MIMA CMth FRAS Polr equtions nd grphs / 6 Adrin Jnnett Objectives The purpose of this presenttion is to cover the following topics:
More informationCone Cluster Labeling for Support Vector Clustering
Cone Cluster Lbeling for Support Vector Clustering Sei-Hyung Lee Deprtment of Computer Science University of Msschusetts Lowell MA 1854, U.S.A. slee@cs.uml.edu Kren M. Dniels Deprtment of Computer Science
More informationMATH 25 CLASS 5 NOTES, SEP
MATH 25 CLASS 5 NOTES, SEP 30 2011 Contents 1. A brief diversion: reltively prime numbers 1 2. Lest common multiples 3 3. Finding ll solutions to x + by = c 4 Quick links to definitions/theorems Euclid
More informationLIMITS AND CONTINUITY
LIMITS AND CONTINUITY Joe McBride/Stone/Gett Imges Air resistnce prevents the velocit of skdiver from incresing indefinitel. The velocit pproches it, clled the terminl velocit. The development of clculus
More information4452 Mathematical Modeling Lecture 4: Lagrange Multipliers
Mth Modeling Lecture 4: Lgrnge Multipliers Pge 4452 Mthemticl Modeling Lecture 4: Lgrnge Multipliers Lgrnge multipliers re high powered mthemticl technique to find the mximum nd minimum of multidimensionl
More informationChapter 2 Sensitivity Analysis: Differential Calculus of Models
Chpter 2 Sensitivity Anlysis: Differentil Clculus of Models Abstrct Models in remote sensing nd in science nd engineering, in generl re, essentilly, functions of discrete model input prmeters, nd/or functionls
More informationa < a+ x < a+2 x < < a+n x = b, n A i n f(x i ) x. i=1 i=1
Mth 33 Volume Stewrt 5.2 Geometry of integrls. In this section, we will lern how to compute volumes using integrls defined by slice nlysis. First, we recll from Clculus I how to compute res. Given the
More informationsuch that the S i cover S, or equivalently S
MATH 55 Triple Integrls Fll 16 1. Definition Given solid in spce, prtition of consists of finite set of solis = { 1,, n } such tht the i cover, or equivlently n i. Furthermore, for ech i, intersects i
More information2 Computing all Intersections of a Set of Segments Line Segment Intersection
15-451/651: Design & Anlysis of Algorithms Novemer 14, 2016 Lecture #21 Sweep-Line nd Segment Intersection lst chnged: Novemer 8, 2017 1 Preliminries The sweep-line prdigm is very powerful lgorithmic design
More informationApproximation by NURBS with free knots
pproximtion by NURBS with free knots M Rndrinrivony G Brunnett echnicl University of Chemnitz Fculty of Computer Science Computer Grphics nd Visuliztion Strße der Ntionen 6 97 Chemnitz Germny Emil: mhrvo@informtiktu-chemnitzde
More informationParallel Square and Cube Computations
Prllel Squre nd Cube Computtions Albert A. Liddicot nd Michel J. Flynn Computer Systems Lbortory, Deprtment of Electricl Engineering Stnford University Gtes Building 5 Serr Mll, Stnford, CA 945, USA liddicot@stnford.edu
More informationEXPONENTIAL & POWER GRAPHS
Eponentil & Power Grphs EXPONENTIAL & POWER GRAPHS www.mthletics.com.u Eponentil EXPONENTIAL & Power & Grphs POWER GRAPHS These re grphs which result from equtions tht re not liner or qudrtic. The eponentil
More informationAn Efficient Divide and Conquer Algorithm for Exact Hazard Free Logic Minimization
An Efficient Divide nd Conquer Algorithm for Exct Hzrd Free Logic Minimiztion J.W.J.M. Rutten, M.R.C.M. Berkelr, C.A.J. vn Eijk, M.A.J. Kolsteren Eindhoven University of Technology Informtion nd Communiction
More informationP(r)dr = probability of generating a random number in the interval dr near r. For this probability idea to make sense we must have
Rndom Numers nd Monte Crlo Methods Rndom Numer Methods The integrtion methods discussed so fr ll re sed upon mking polynomil pproximtions to the integrnd. Another clss of numericl methods relies upon using
More informationUSING HOUGH TRANSFORM IN LINE EXTRACTION
Stylinidis, Efstrtios USING HOUGH TRANSFORM IN LINE EXTRACTION Efstrtios STYLIANIDIS, Petros PATIAS The Aristotle University of Thessloniki, Deprtment of Cdstre Photogrmmetry nd Crtogrphy Univ. Box 473,
More information)
Chpter Five /SOLUTIONS Since the speed ws between nd mph during this five minute period, the fuel efficienc during this period is between 5 mpg nd 8 mpg. So the fuel used during this period is between
More informationSpring 2018 Midterm Exam 1 March 1, You may not use any books, notes, or electronic devices during this exam.
15-112 Spring 2018 Midterm Exm 1 Mrch 1, 2018 Nme: Andrew ID: Recittion Section: You my not use ny books, notes, or electronic devices during this exm. You my not sk questions bout the exm except for lnguge
More informationPreserving Constraints for Aggregation Relationship Type Update in XML Document
Preserving Constrints for Aggregtion Reltionship Type Updte in XML Document Eric Prdede 1, J. Wenny Rhyu 1, nd Dvid Tnir 2 1 Deprtment of Computer Science nd Computer Engineering, L Trobe University, Bundoor
More informationHeuristics for Thelen s Prime Implicant Method 1
S C H E D A E I N F O R M A T I C A E VOLUME 14 2005 Heuristics for Thelen s Prime Implicnt Method 1 Jcek Biegnowski, Andrei Krtkevich Institute of Computer Engineering nd Electronics, ul. Podgórn 50,
More informationReplicating Web Applications On-Demand
Replicting Web Applictions On-Demnd Swminthn Sivsubrmnin Guillume Pierre Mrten vn Steen Dept. of Computer Science, Vrije Universiteit, Amsterdm {swmi,gpierre,steen}@cs.vu.nl Abstrct Mny Web-bsed commercil
More informationNearest Keyword Set Search in Multi-dimensional Datasets
Nerest Keyword Set Serch in Multi-dimensionl Dtsets Vishwkrm Singh Deprtment of Computer Science University of Cliforni Snt Brbr, USA Emil: vsingh014@gmil.com Ambuj K. Singh Deprtment of Computer Science
More informationIMAGE QUALITY OPTIMIZATION BASED ON WAVELET FILTER DESIGN AND WAVELET DECOMPOSITION IN JPEG2000. Do Quan and Yo-Sung Ho
IMAGE QUALITY OPTIMIZATIO BASED O WAVELET FILTER DESIG AD WAVELET DECOMPOSITIO I JPEG2000 Do Qun nd Yo-Sung Ho School of Informtion & Mechtronics Gwngju Institute of Science nd Technology (GIST) 26 Cheomdn-gwgiro
More informationTree Structured Symmetrical Systems of Linear Equations and their Graphical Solution
Proceedings of the World Congress on Engineering nd Computer Science 4 Vol I WCECS 4, -4 October, 4, Sn Frncisco, USA Tree Structured Symmetricl Systems of Liner Equtions nd their Grphicl Solution Jime
More informationSection 10.4 Hyperbolas
66 Section 10.4 Hyperbols Objective : Definition of hyperbol & hyperbols centered t (0, 0). The third type of conic we will study is the hyperbol. It is defined in the sme mnner tht we defined the prbol
More informationMemory-Optimized Software Synthesis from Dataflow Program Graphs withlargesizedatasamples
EURSIP Journl on pplied Signl Processing 2003:6, 54 529 c 2003 Hindwi Publishing orportion Memory-Optimized Softwre Synthesis from tflow Progrm Grphs withlrgesizetsmples Hyunok Oh The School of Electricl
More informationEfficient K-NN Search in Polyphonic Music Databases Using a Lower Bounding Mechanism
Efficient K-NN Serch in Polyphonic Music Dtses Using Lower Bounding Mechnism Ning-Hn Liu Deprtment of Computer Science Ntionl Tsing Hu University Hsinchu,Tiwn 300, R.O.C 886-3-575679 nhliou@yhoo.com.tw
More informationINTRODUCTION TO SIMPLICIAL COMPLEXES
INTRODUCTION TO SIMPLICIAL COMPLEXES CASEY KELLEHER AND ALESSANDRA PANTANO 0.1. Introduction. In this ctivity set we re going to introduce notion from Algebric Topology clled simplicil homology. The min
More informationMA1008. Calculus and Linear Algebra for Engineers. Course Notes for Section B. Stephen Wills. Department of Mathematics. University College Cork
MA1008 Clculus nd Liner Algebr for Engineers Course Notes for Section B Stephen Wills Deprtment of Mthemtics University College Cork s.wills@ucc.ie http://euclid.ucc.ie/pges/stff/wills/teching/m1008/ma1008.html
More informationAn Algorithm for Enumerating All Maximal Tree Patterns Without Duplication Using Succinct Data Structure
, Mrch 12-14, 2014, Hong Kong An Algorithm for Enumerting All Mximl Tree Ptterns Without Dupliction Using Succinct Dt Structure Yuko ITOKAWA, Tomoyuki UCHIDA nd Motoki SANO Astrct In order to extrct structured
More information1.1 Lines AP Calculus
. Lines AP Clculus. LINES Notecrds from Section.: Rules for Rounding Round or Truncte ll finl nswers to 3 deciml plces. Do NOT round before ou rech our finl nswer. Much of Clculus focuses on the concept
More informationFig.25: the Role of LEX
The Lnguge for Specifying Lexicl Anlyzer We shll now study how to uild lexicl nlyzer from specifiction of tokens in the form of list of regulr expressions The discussion centers round the design of n existing
More informationBefore We Begin. Introduction to Spatial Domain Filtering. Introduction to Digital Image Processing. Overview (1): Administrative Details (1):
Overview (): Before We Begin Administrtive detils Review some questions to consider Winter 2006 Imge Enhncement in the Sptil Domin: Bsics of Sptil Filtering, Smoothing Sptil Filters, Order Sttistics Filters
More informationDigital Design. Chapter 6: Optimizations and Tradeoffs
Digitl Design Chpter 6: Optimiztions nd Trdeoffs Slides to ccompny the tetbook Digitl Design, with RTL Design, VHDL, nd Verilog, 2nd Edition, by Frnk Vhid, John Wiley nd Sons Publishers, 2. http://www.ddvhid.com
More informationOn the Detection of Step Edges in Algorithms Based on Gradient Vector Analysis
On the Detection of Step Edges in Algorithms Bsed on Grdient Vector Anlysis A. Lrr6, E. Montseny Computer Engineering Dept. Universitt Rovir i Virgili Crreter de Slou sin 43006 Trrgon, Spin Emil: lrre@etse.urv.es
More informationRay surface intersections
Ry surfce intersections Some primitives Finite primitives: polygons spheres, cylinders, cones prts of generl qudrics Infinite primitives: plnes infinite cylinders nd cones generl qudrics A finite primitive
More informationAlgorithm Design (5) Text Search
Algorithm Design (5) Text Serch Tkshi Chikym School of Engineering The University of Tokyo Text Serch Find sustring tht mtches the given key string in text dt of lrge mount Key string: chr x[m] Text Dt:
More informationInference of node replacement graph grammars
Glley Proof 22/6/27; :6 File: id293.tex; BOKCTP/Hin p. Intelligent Dt Anlysis (27) 24 IOS Press Inference of node replcement grph grmmrs Jcek P. Kukluk, Lwrence B. Holder nd Dine J. Cook Deprtment of Computer
More informationUNIT 11. Query Optimization
UNIT Query Optimiztion Contents Introduction to Query Optimiztion 2 The Optimiztion Process: An Overview 3 Optimiztion in System R 4 Optimiztion in INGRES 5 Implementing the Join Opertors Wei-Png Yng,
More informationGENERATING ORTHOIMAGES FOR CLOSE-RANGE OBJECTS BY AUTOMATICALLY DETECTING BREAKLINES
GENEATING OTHOIMAGES FO CLOSE-ANGE OBJECTS BY AUTOMATICALLY DETECTING BEAKLINES Efstrtios Stylinidis 1, Lzros Sechidis 1, Petros Ptis 1, Spiros Sptls 2 Aristotle University of Thessloniki 1 Deprtment of
More informationAML710 CAD LECTURE 16 SURFACES. 1. Analytical Surfaces 2. Synthetic Surfaces
AML7 CAD LECTURE 6 SURFACES. Anlticl Surfces. Snthetic Surfces Surfce Representtion From CAD/CAM point of view surfces re s importnt s curves nd solids. We need to hve n ide of curves for surfce cretion.
More informationStack. A list whose end points are pointed by top and bottom
4. Stck Stck A list whose end points re pointed by top nd bottom Insertion nd deletion tke plce t the top (cf: Wht is the difference between Stck nd Arry?) Bottom is constnt, but top grows nd shrinks!
More informationCS201 Discussion 10 DRAWTREE + TRIES
CS201 Discussion 10 DRAWTREE + TRIES DrwTree First instinct: recursion As very generic structure, we could tckle this problem s follows: drw(): Find the root drw(root) drw(root): Write the line for the
More informationA Formalism for Functionality Preserving System Level Transformations
A Formlism for Functionlity Preserving System Level Trnsformtions Smr Abdi Dniel Gjski Center for Embedded Computer Systems UC Irvine Center for Embedded Computer Systems UC Irvine Irvine, CA 92697 Irvine,
More informationSolutions to Math 41 Final Exam December 12, 2011
Solutions to Mth Finl Em December,. ( points) Find ech of the following its, with justifiction. If there is n infinite it, then eplin whether it is or. ( ) / ln() () (5 points) First we compute the it:
More informationCross-Supervised Synthesis of Web-Crawlers
Cross-Supervised Synthesis of Web-Crwlers Adi Omri Technion omri@cs.technion.c.il Shron Shohm Acdemic College of Tel Aviv Yffo shron.shohm@gmil.com Ern Yhv Technion yhve@cs.technion.c.il ABSTRACT A web-crwler
More informationSAPPER: Subgraph Indexing and Approximate Matching in Large Graphs
SAPPER: Sugrph Indexing nd Approximte Mtching in Lrge Grphs Shijie Zhng, Jiong Yng, Wei Jin EECS Dept., Cse Western Reserve University, {shijie.zhng, jiong.yng, wei.jin}@cse.edu ABSTRACT With the emergence
More informationDynamic Programming. Andreas Klappenecker. [partially based on slides by Prof. Welch] Monday, September 24, 2012
Dynmic Progrmming Andres Klppenecker [prtilly bsed on slides by Prof. Welch] 1 Dynmic Progrmming Optiml substructure An optiml solution to the problem contins within it optiml solutions to subproblems.
More informationThe Basic Properties of the Integral
The Bsic Properties of the Integrl When we compute the derivtive of complicted function, like + sin, we usull use differentition rules, like d [f()+g()] d f()+ d g(), to reduce the computtion d d d to
More informationAn Efficient and Effective Case Classification Method Based On Slicing
An Efficient nd Effective Cse Clssifiction Method Bsed On Slicing An Efficient nd Effective Cse Clssifiction Method Bsed On Slicing Omr A. A. Shib, Md. Nsir Sulimn, Ali Mmt nd Ftimh Ahmd Fculty of Computer
More informationcell 1 cell 2 cell 3 cell 4 a a
Frequency Assignment for Cellulr Mobile Systems Using Constrint Stisfction Techniques Mkoto Yokooy nd Ktsutoshi Hirymz y NTT Communiction Science Lbortories -4 Hikridi, Seik-cho, Sorku-gun, Kyoto 69-037
More informationVirtual Network Embedding For Evolving Networks
This full text pper ws peer reviewed t the direction of IEEE Communictions Society subject mtter experts for publiction in the IEEE Globecom proceedings. Virtul Network Embedding For Evolving Networks
More informationAlignment of Long Sequences. BMI/CS Spring 2012 Colin Dewey
Alignment of Long Sequences BMI/CS 776 www.biostt.wisc.edu/bmi776/ Spring 2012 Colin Dewey cdewey@biostt.wisc.edu Gols for Lecture the key concepts to understnd re the following how lrge-scle lignment
More informationA HYDRAULIC SIMULATOR FOR AN EXCAVATOR
P-06 Proceedings of the 7th JFPS Interntionl Symposium on Fluid Power TOYAMA 008 September 5-8 008 A HYDRAULIC SIMULATOR FOR AN EXCAVATOR Soon-Kwng Kwon* Je-Jun Kim* Young-Mn Jung* Chn-Se Jung* Chng-Don
More informationOptimal aeroacoustic shape design using approximation modeling
Center for Turbulence Reserch Annul Reserch Briefs 22 21 Optiml erocoustic shpe design using pproximtion modeling By Alison L. Mrsden, Meng Wng AND Petros Koumoutskos 1. Introduction Reduction of noise
More informationImproper Integrals. October 4, 2017
Improper Integrls October 4, 7 Introduction We hve seen how to clculte definite integrl when the it is rel number. However, there re times when we re interested to compute the integrl sy for emple 3. Here
More informationEpson Projector Content Manager Operation Guide
Epson Projector Content Mnger Opertion Guide Contents 2 Introduction to the Epson Projector Content Mnger Softwre 3 Epson Projector Content Mnger Fetures... 4 Setting Up the Softwre for the First Time
More informationIf you are at the university, either physically or via the VPN, you can download the chapters of this book as PDFs.
Lecture 5 Wlks, Trils, Pths nd Connectedness Reding: Some of the mteril in this lecture comes from Section 1.2 of Dieter Jungnickel (2008), Grphs, Networks nd Algorithms, 3rd edition, which is ville online
More informationChapter Spline Method of Interpolation More Examples Electrical Engineering
Chpter. Spline Method of Interpoltion More Exmples Electricl Engineering Exmple Thermistors re used to mesure the temperture of bodies. Thermistors re bsed on mterils chnge in resistnce with temperture.
More informationCOMP 423 lecture 11 Jan. 28, 2008
COMP 423 lecture 11 Jn. 28, 2008 Up to now, we hve looked t how some symols in n lphet occur more frequently thn others nd how we cn sve its y using code such tht the codewords for more frequently occuring
More informationPointwise convergence need not behave well with respect to standard properties such as continuity.
Chpter 3 Uniform Convergence Lecture 9 Sequences of functions re of gret importnce in mny res of pure nd pplied mthemtics, nd their properties cn often be studied in the context of metric spces, s in Exmples
More informationWhat are suffix trees?
Suffix Trees 1 Wht re suffix trees? Allow lgorithm designers to store very lrge mount of informtion out strings while still keeping within liner spce Allow users to serch for new strings in the originl
More informationOverview. Network characteristics. Network architecture. Data dissemination. Network characteristics (cont d) Mobile computing and databases
Overview Mobile computing nd dtbses Generl issues in mobile dt mngement Dt dissemintion Dt consistency Loction dependent queries Interfces Detils of brodcst disks thlis klfigopoulos Network rchitecture
More informationSeamless Astronomy Alyssa A. Goodman
Semless Astronomy Alyss A. Goodmn Hrvrd-Smithsonin Center for Astrophysics Inititive in Innovtive Computing @ Hrvrd Collbortors Hrvrd-Smithsonin Center for Astrophysics & SEAS: Alberto Accomzzi, Eli Bressert,
More informationTilt-Sensing with Kionix MEMS Accelerometers
Tilt-Sensing with Kionix MEMS Accelerometers Introduction Tilt/Inclintion sensing is common ppliction for low-g ccelerometers. This ppliction note describes how to use Kionix MEMS low-g ccelerometers to
More informationControl-Flow Analysis and Loop Detection
! Control-Flow Anlysis nd Loop Detection!Lst time! PRE!Tody! Control-flow nlysis! Loops! Identifying loops using domintors! Reducibility! Using loop identifiction to identify induction vribles CS553 Lecture
More informationPerformance enhancement of IEEE DCF using novel backoff algorithm
Kuo et l. EURASIP Journl on Wireless Communictions nd Networking 212, 212:274 http://jis.eursipjournls.com/content/212/1/274 RESEARCH Open Access Performnce enhncement of IEEE 82.11 using novel ckoff lgorithm
More informationIn the last lecture, we discussed how valid tokens may be specified by regular expressions.
LECTURE 5 Scnning SYNTAX ANALYSIS We know from our previous lectures tht the process of verifying the syntx of the progrm is performed in two stges: Scnning: Identifying nd verifying tokens in progrm.
More informationOPERATION MANUAL. DIGIFORCE 9307 PROFINET Integration into TIA Portal
OPERATION MANUAL DIGIFORCE 9307 PROFINET Integrtion into TIA Portl Mnufcturer: 2018 burster präzisionsmesstechnik gmbh & co kg burster präzisionsmesstechnik gmbh & co kg Alle Rechte vorbehlten Tlstrße
More information9 Graph Cutting Procedures
9 Grph Cutting Procedures Lst clss we begn looking t how to embed rbitrry metrics into distributions of trees, nd proved the following theorem due to Brtl (1996): Theorem 9.1 (Brtl (1996)) Given metric
More informationCSCI 446: Artificial Intelligence
CSCI 446: Artificil Intelligence Serch Instructor: Michele Vn Dyne [These slides were creted by Dn Klein nd Pieter Abbeel for CS188 Intro to AI t UC Berkeley. All CS188 mterils re vilble t http://i.berkeley.edu.]
More informationPerformance analysis of QoS mechanisms in IP networks
University of Wollongong Reserch Online Fculty of Informtics - Ppers (Archive) Fculty of Engineering nd Informtion Sciences 2000 Performnce nlysis of QoS mechnisms in IP networks D. Ji University of Wollongong
More informationDYNAMIC DISCRETIZATION: A COMBINATION APPROACH
DYNAMIC DISCRETIZATION: A COMBINATION APPROACH FAN MIN, QIHE LIU, HONGBIN CAI, AND ZHONGJIAN BAI School of Computer Science nd Engineering, University of Electronic Science nd Technology of Chin, Chengdu
More informationGraph Theory and DNA Nanostructures. Laura Beaudin, Jo Ellis-Monaghan*, Natasha Jonoska, David Miller, and Greta Pangborn
Grph Theory nd DNA Nnostructures Lur Beudin, Jo Ellis-Monghn*, Ntsh Jonosk, Dvid Miller, nd Gret Pngborn A grph is set of vertices (dots) with edges (lines) connecting them. 1 2 4 6 5 3 A grph F A B C
More information