Classification of binary vectors by using DSC distance to minimize stochastic complexity
|
|
- Pierce Gaines
- 5 years ago
- Views:
Transcription
1 Patter Recogitio Letters 24 (2003) Classificatio of biary vectors by usig DSC distace to miimize stochastic complexity Pasi Fr ati *, Matao Xu, Ismo K arkk aie Departmet of Computer Sciece, Uiversity of Joesuu, P.O. Box 111, Fi Joesuu, Filad Received 17 October 2001; received i revised form 28 February 2002 Abstract Stochastic complexity (SC) has bee employed as a cost fuctio for solvig biary clusterig problem usig Shao code legth (CL distace) as the distace fuctio. The CL distace, however, is defied for a give static clusterig oly, ad it does ot take ito accout of the chages i the class distributio durig the clusterig process. We propose a ew DSC distace fuctio, which is derived directly from the differece of the cost fuctio value before ad after the classificatio. The effect of the ew distace fuctio is demostrated by implemetig it with two clusterig algorithms. Ó 2002 Elsevier Sciece B.V. All rights reserved. Keywords: Clusterig; Stochastic complexity; Algorithms; Distace fuctios 1. Itroductio Biary vector classificatio has bee widely used i DNA computig ad huma chromosome study, ad i solvig taxoomy problems from biomedical area. Statistical models are usually applied to describe ad solve predictio ad taxoomy problems. For example, Rissae (1987, 1996) has itroduced a model kow as stochastic complexity (SC), which is a extesible explaatio for Shao iformatio theory (Kotkae et al., 1999). To be a cost fuctio of classificatio, SC eeds to be approximated by a simple model (Rissae, * Correspodig author. Tel.: ; fax: address: frati@cs.joesuu.fi (P. Fr ati). 1987). Gylleberg et al. (1994, 1997, 2000) has give a simple ad practical approximatio of SC for biary vector classificatios. Thereafter, SC has bee employed as a geeric evaluatio fuctio i solvig biary clusterig problems as follows. The clusterig problem is first formulated as a optimizatio problem. Approximatio solutios are the foud for every reasoable umber of groups. SC is applied for measurig the goodess of the various clusterig results. Idividual clusterig ca be geerated usig ay algorithms such as the geeralized Lloyd algorithm (GLA) (Lide et al., 1980), ad the radomized local search (RLS) Fr ati ad Kivij arvi, A better idea is to itegrate the SC cost fuctio directly i the clusterig algorithm as doe i Gylleberg et al. (1997) ad Fr ati et al. (2000). The vector-to-cluster distace for the classificatio /03/$ - see frot matter Ó 2002 Elsevier Sciece B.V. All rights reserved. PII: S (02)
2 66 P. Fr ati et al. / Patter Recogitio Letters 24 (2003) of the vectors must the be re-defied correspodigly. Euclidea distace (L 2 -orm) provides the optimal classificatio of the data vectors for the miimizatio of the MSE, but ot for the SC. The optimal classificatio for SC is give by the Shao code legth (CL) fuctio (Gylleberg et al., 1997). It represets the etropy of the biary vector whe coded by the probability model of the particular cluster. Surprisigly, the CL distace itroduces a ew problem that ever arises with the L 2 -distace. This is illustrated i Fig. 1, i which we classify the black poit accordig the two existig clusters. The probability distributio of the leftmost cluster idicates the poit belogs to this class with a low probability. Nevertheless, the probability distributio of the rightmost cluster have zero variace i the horizotal dimesio ðr x ¼ 0Þ resultig i zero probability ad ifiite etropy. As a cosequece, the poit will be classified to the leftmost cluster. This ifiite etropy problem happes ofte i the classificatio of multi-dimesioal biary data vectors. It has therefore bee ecessary to make modificatios to the existig clusterig algorithms whe SC has bee applied as a cost fuctio. Previously, the problem has bee solved i Gylleberg et al. (1997) ad Fr ati et al. (2000) by applyig the clusterig algorithms first usig the sub-optimal but less problematic L 2 -distace. The CL distace is the applied i the last stage of the algorithm whe the global clusterig structure has already settled dow ad oly fie-tuig of the solutio takes place. The drawback of this approach is that similar patch should be made for every clusterig algorithm that is to be applied with the SC. I this paper, we propose a more geeral solutio to the ifiity problem by proposig a ew Fig. 1. Illustrative example of the problem i the CL distace. DSC distace fuctio. The distace fuctio is derived directly from the differece of the cost fuctio value before ad after the classificatio. It therefore implicitly takes ito accout the chage i the class distributio caused by the re-classificatio of the data vector, ad i this way, avoids the ifiity problem. The DSC is geeral i the sese that it applies to ay clusterig algorithm ad o more patches are therefore eeded. The effect of the ew distace fuctio is demostrated by implemetig it with two clusterig algorithms. The rest of the paper is orgaized as follows. I Sectio 2, we defie the clusterig problem of biary vectors, ad give the simplified formalizatio of the SC. The SC fuctio is the applied withi two clusterig algorithms as the cost fuctio, ad the CL distace is employed i the RLS ad GLA algorithms as a practical vector-to-cluster distace. I Sectio 3, we itroduce the ew DSC fuctio derived from the SC differece of the old ad ew classificatio whe a data vector is moved from oe class to aother. I Sectio 4, we make performace comparisos of the differet variats icludig the RLS ad GLA algorithms, ad the DSC, CL ad L 2 distace fuctios. 2. Clusterig by miimizig SC We use the followig otatios: N: umber of data vectors, M: umber of groups, D: dimesio of vectors, X: set of N data objects X ¼fx 1 ; x 2 ;...; x N g, P: partitio idices of x i : P ¼fp i ji ¼ 1;...; Ng, C: set of cluster cetroids C ¼fc j jj ¼ 1;...; Mg. The goal of the clusterig is to partitio a give set of N data vectors ito a umber of groups so that a give cost fuctio is miimized. I the clusterig process, we must solve both the umber of clusters (M) ad their locatio (c i ). The clusterig result is described by the partitio (P) of the data set by givig for each vector (x i ) the cluster idex (p i ) of the group, which it belogs to. We
3 P. Fr ati et al. / Patter Recogitio Letters 24 (2003) cosider a set of d-dimesioal biary data vectors Stochastic complexity SC ca be applied to the clusterig by fidig the miimum descriptio of the data via usig a clusterig model. SC measures the iformatio cotet of the data, ad it is defied as the shortest possible code legth for the data obtaiable by usig a set of class distributios. SC icludes both the model parameters ad the codig of the data i the measuremet. Suppose that we have classified the data vectors ito M groups described by the partitio of the data. The model of the class j ca the be described by the probability distributio withi the class i each dimesio: c ij ¼ ij = j ð1þ where j is the umber of biary vectors i the class j, ad ij is the umber of vectors havig the ith coordiate value 1. The probability vector c j of the class j is also the cetroid (average vector) of the cluster. The simplified approximatio of the SC fuctio i Gylleberg et al. (1997) ca be described usig the class distributio models as: SC ¼ XM X d j h ij þ XM j log j j N j¼1 þ d 2 X M j 1 j¼1 log maxð1; j Þ ð2þ where h measures the etropy of a biary distributio hðpþ ¼ p logðpþ ð1 pþlogð1 pþ ð3þ Sice every vector is classified to some group, it is kow that P j ¼ N. Moreover, N log N i the middle term is a costat ad, therefore, the equatio ca be simplified as: SC XM X d j h ij þ XM j log j þ d j¼1 j 2 j¼1 However, the simplified Eq. (4) above could make SC egative. The first part of the SC fuctio measures the itra-class iformatio as the code legth whe every data vector is coded accordig to the class probability model. The code legth is calculated by multiplyig the umber of vectors i each cluster ( j ) by the average etropy (h) ofthe cluster. The secod part measures the iter-class iformatio as the code legth of the partitio. It ca be calculated by the umber of vectors i each cluster ( j ) multiplied by the average etropy of the correspodig cluster idex. The third part measures the iformatio of the model as the code legth of the class distributio whe described by a series umbers betwee ½1:: j Š Clusterig algorithm The SC ca be applied for the clusterig problem as follows. We first fid approximatio solutios for every reasoable umber of groups usig ay clusterig algorithm. The solutios are the evaluated, ad the oe that miimizes the SC is the fial result of the clusterig. This search strategy ca use ay clusterig algorithm to fid the idividual solutios. I the followig, we recall two clusterig algorithms: the GLA by Lide et al. (1980), ad the RLS by Fr ati ad Kivij arvi (2000). The pseudocode of the GLA is show i Fig. 2. The algorithm takes ay iitial solutio (here the partitio P) as a iput, ad iteratively fie-tues the solutio by repeatig two operatios i tur. The first operatio calculates the cetroids of the clusters, ad the secod operatio re-partitio the data vectors accordig to the ew set of cetroids. XM j¼1 log maxð1; j Þ ð4þ Fig. 2. Pseudocode for the GLA.
4 68 P. Fr ati et al. / Patter Recogitio Letters 24 (2003) Fig. 3. Pseudocode for the RLS. The algorithm is iterated util o more improvemet appears i the solutio. The method is simple to implemet, ad has bee widely used for the clusterig problem as such, or itegrated with more complicated methods. The pseudocode of the RLS is show i Fig. 3. The method takes ay iitial solutio, which is the improved by a sequece of operatios. At each iteratio phase, the algorithm creates a ew cadidate solutio by makig a small chage to the curret clusterig structure. First, a radomly chose cluster cetroid (c j ) is replaced by a radomly chose data vector (x i ). This moves the cluster locatio to aother part of the vector space. The partitio is the adjusted by a local repartitio operatio, which cosists of the two steps as show i Fig. 4. I the first step, the old cluster is removed by re-partitioig its data vectors to other clusters. I the secod step, the ewly created cluster is populated by attractig data vectors from the eighborig clusters. The modified clusterig is fie-tued by the applicatio of the GLA. The ew cadidate solutio is the evaluated ad accepted oly if it improves the previous solutio. Fig. 4. Pseudocode for the local repartitio operatio. Otherwise, the cadidate solutio is discarded ad the previous solutio remais as the startig poit for the ext iteratio. The GLA ad the RLS are both applicable for the clusterig task ad also rather simple to implemet. The RLS is less sesitive to the iitializatio because it is capable of makig global chages i the clusterig structure (by radom swappig of the clusters), ad therefore, correct the icorrect settlemet of the iitial clusterig. If the GLA is to be used, it should be repeated several times i order to reduce the depedecy o the iitializatio Shao code-legth distace The clusterig algorithms employ a distace fuctio d, which measures the vector-to-cluster distace, ad is used for the classificatio of the vectors durig the clusterig process. Usually the distace fuctio is defied as the Euclidea distace (L 2 -orm) betwee the data vector x i ad the particular cluster cetroid c j : vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ux d d E ðx i ; c j Þ¼t kx ik c jk k 2 ð5þ k¼1 This gives optimal classificatio for the miimizatio of the MSE but ot for the SC. The optimal classificatio for the SC is give by the Shao code-legth fuctio CLðx i ; c j Þ i Gylleberg et al. (1997). d CL ðx i ; c j Þ¼ Xd ð1 x i Þ logð1 c ij Þþx i log c ij log j ð6þ N It measures the code legth whe the data vector (the summatio term i the equatio) ad its class idex (secod term i the equatio) are coded usig the give model. I priciple, the CL distace is well defied, but i practice, it has a fudametal problem i its defiitio, which will be explaied by the followig example. Cosider a sigle cluster c 1 cosistig of the followig three data vectors: x 1 ¼ð0; 0; 0Þ, x 2 ¼ ð0; 1; 0Þ ad x 3 ¼ð1; 0; 0Þ. The correspodig class
5 P. Fr ati et al. / Patter Recogitio Letters 24 (2003) probability distributio of the cluster is c 1 ¼ð0:33; 0:33; 0:00Þ. The distaces of the vectors ca ow be calculated usig the CL distace: d CL ðx 1 ; c 1 Þ¼0:58 þ 0:58 þ 0:00 ¼ 1:17 d CL ðx 2 ; c 1 Þ¼0:58 þ 1:58 þ 0:00 ¼ 2:17 d CL ðx 3 ; c 1 Þ¼1:58 þ 0:58 þ 0:00 ¼ 2:17 The secod term of Eq. (6) is omitted i this example for simplicity. Let us the cosider a fourth vector x 4 ¼ð0; 0; 1Þ, which is equally close to the cluster accordig to the Euclidea distace, but the CL distace gives the followig result: d CL ðx 4 ; c 1 Þ¼0:58 þ 0:58 þ1¼udefied The problem ca appear whe there is a uiform bit distributio i ay dimesio, ad the data vector has differet value i the same positio. The homogeous bit distributio idicates that there is o ucertaity ad the etropy of the cotradictig value would therefore approach ifiite. This is a serious flaw especially i the local re-partitio procedure of the RLS. It creates ew clusters startig from a sigular cluster, which evidetly has uiform bit distributio. As a cosequece, o other data vectors (except equal oes) ca ever be classified to this cluster. The problem of the CL distace is that eve though it measures the ucertaity of the classificatio, it does ot take ito accout the ucertaity of the model itself. I other words, the bit distributio of the class model is ideed homogeeous, but the model is oly a approximatio ad subject to chage durig the clusterig process. Zero-probability is therefore ot a feasible approximatio of the classificatio. The ifiite values could be avoided by prevetig the cetroids to take values 0 ad 1. This ca be achieved, for example, give biary data vector x, by takig the cetroid values to be mea vector of x ad c j. If some coordiate of c j equals to 0 or 1 value, the umber of vectors i jth cluster, j ca be take as a parameter of CL distace fuctio. Obviously, c j ca be replaced with a ew vector, which is the cetroid of jth cluster after vector x is put ito cluster j. c ij ¼ jc ij þ x i j þ 1 c ij 6¼ x i ð7þ If c j ¼ x, CL distace is adopted as log 2 ðnþ. Aother coditio o CL distace value is cosidered as follows: x i log c ij þð1 x i Þ logð1 c ij Þ¼0 c ij ¼ x i ; c j 6¼ x ð8þ This patch, however, does ot remove the problem itself as it merely assigs a low probability istead of a zero value. Additioal modificatios have therefore bee ecessary for the clusterig algorithms so that the CL distace could have bee used properly i the GLA ad i the RLS. For example, i the algorithms preseted i Gylleberg et al. (1997) ad Fr ati et al. (2000) the CL distace is applied oly i the last step of the clusterig process whe the global clusterig structure has already settled dow, ad oly fie-tuig of the solutio takes place. The problem of this approach is that it is ot trivial to determie the stage of the clusterig process, whe it would be safe eough to start to use the CL distace. To sum up, the problem with the CL distace is fudametal i its ature. It is therefore better to fix it tha to fid patch for every clusterig algorithm that is to be applied. 3. DSC distace fuctio We itroduce a ew vector-to-cluster distace fuctio deoted as DSC distace. It is based o a desig paradigm, i which the distace fuctio is derived directly from the differece of the cost fuctio value before ad after the classificatio of a data vector. The mai advatage of this desig philosophy is that it implicitly takes ito accout of the chages i the clusterig model caused by the classificatio. It is also geeral i the sese that it does ot deped o the chose clusterig algorithm ad should therefore be applicable with ay distace-based clusterig method. The DSC distace fuctio is always defied relative to a give model. We ca therefore assume that we have a model, for which we ca calculate the SC value. If we the cosider the distace
6 70 P. Fr ati et al. / Patter Recogitio Letters 24 (2003) calculatio as a movemet of the data vector from oe group to aother, we ca defie the distace fuctio as the differece i the SC of the clusterig before ad after the movemet of the data vector. Give two classes j 1, j 2 ad a biary vector x, which we cosider to move from the class j 1 to class j 2, the SC fuctio i (2) value after the movemet is: SC XM X d j j6¼j 1 ;j 2 þ d 2 X M j6¼j 1 ;j 2 h ij j þ X j6¼j 1 ;j 2 ð j log 2ð j ÞÞ log maxð1; j Þ ð j1 1Þ logð j1 1Þ ð j2 þ 1Þ logð j2 þ 1Þ þ Xd 1Þh x i ð j1 j1 1 þð j2 þ 1Þh þ x i j2 þ 1 þ d 2 ðlog maxð1; j 1 1Þ þ log maxð1; j2 þ 1ÞÞ þ N log N ð9þ We ca the calculate the differece betwee the SC fuctio values of the old clusterig (before the movemet) ad the ew oe (after the movemet) as: SC-diffðx; j 1 ; j 2 Þ¼ Xd j1 1 h x i j1 1 j1 h j1 þð j2 þ 1Þh þ x i j2 þ 1 j2 h þð j1 d=2þ j2 log j1 þð j2 d=2þ log j2 ð j2 þ 1 d=2þ logð j2 þ 1Þ ð j1 1 d=2þ logð j1 1Þ j1 > 1 SC-diffðx; j 1 ; j 2 Þ¼ Xd j2 þ 1 h x i j2 1 j2 h j2 þð j2 d=2þ log j2 þðd=2 j2 1Þ logð j2 þ 1Þ j1 ¼ 1 ð10þ The SC-diff takes zero value if j 1 ¼ j 2. Negative values are obtaied whe the movemet of the vector improves the solutio, ad positive values otherwise. The SC-diff could ow be applied as such i the cases whe we re-classify a vector i a existig solutio. I the SC-diff fuctio we assume that the give vector is already classified ito some class. I geeral, however, this is ot the case but we must be able to defie a more geeral distace fuctio that depeds oly o the vector x i ad o the cadidate cluster c j. For example, i the repartitio procedure of the RLS algorithm, we classify vectors whose previous class has bee removed. More geeral DSC fuctio ca be derived from (10) as follows. The classificatio ca be cosidered as a twostep procedure, i which we first remove the vector x i from the class j 1 ad the add it to the class j 2 For a give vector x i the cost of the removal is costat. This meas that the parameters N, j1, ij1 are fixed i the classificatio, ad as a cosequece, we ca cosider oly the cost of addig the vector i class j 2 ad igore the removal part i the formula. Thus, the DSC ðj 1 6¼ j 2 Þ ca be defied merely as the cost of the additio DSCðx; C j2 Þ¼ Xd j2 þ 1 h þ x i j2 þ 1 j2 h j2 þð j2 d=2þ log j2 þðd=2 j2 1Þ logð j2 þ 1Þ þ log N ð11þ
7 P. Fr ati et al. / Patter Recogitio Letters 24 (2003) This gives the same result as the SC-diff with the differece of a costat. The oly exceptio is whe we measure the distace of x i to the cluster, i which it is already icluded ðj 1 ¼ j 2 Þ. I this case, we should use ð j1 1Þ as the class size istead of j1 because the class size does ot icrease due to the classificatio. Thus, if the previous classificatio is kow, we should apply the followig equatio for this special case: DSCðx; C j2 Þ¼ Xd j1 h j1 ð j1 1Þh x i j1 1 Fig. 5. Clusterig results by the RLS algorithm for DNA-1. þðd=2 j1 Þ log j1 þð j1 1 d=2þ logð j1 1Þ þ log N j 1 ¼ j 2 ; j1 > 1 DSCðx; C j2 Þ¼log N j 1 ¼ j 2 ; j1 ¼ 1 ð12þ Hece, DSC distace as i Eq. (11) is applicable as vector-to-cluster distace i all cases, although it uderestimates the distace i the case whe the vector is already icluded i the class. The special case of (12) should therefore be used whe applicable to give more exact value. 4. Test results We use three biary data sets to test the ew method: DNA-1, DNA-2, Normal. The features of the first two sets (DNA-1 ad DNA-2) were extracted from aalysis of DNA samples of fishes (presece or absece of give DNA fragmet) i biological research experimets. There are dimesioal biary vectors i (DNA-1) ad dimesioal vectors i DNA-2. The third set (Normal) was artificially created by geeratig 265 biary vectors ito 10-dimesioal vector space with 12 clusters. We study first the DNA-1 ad DNA-2 sets whe clustered usig the RLS ad GLA methods. The RLS was performed 80 iteratios. The results i Figs. 5 8 show that the DSC distace ad L 2 -distace Fig. 6. Clusterig results by the GLA algorithm for DNA-1. Fig. 7. Clusterig results by the RLS algorithm for DNA-2.
8 72 P. Fr ati et al. / Patter Recogitio Letters 24 (2003) Table 1 summarizes the clusterig ad classificatio results for the Normal data set. It is the oly data set for which the real classificatio is kow, ad thus, classificatio rate could be calculated. The results show that employig RLS by DSC distace gives the best performaces both i terms of best clusterig result (smallest SC values), ad the highest classificatio rate. The RLS algorithm foud the correct umber of clusters also with the L 2 -distace ad CL distace but the correspodig classificatio rates were smaller. The GLA, however, was able to fid the correct result (12 clusters) oly by usig the DSC distace. Fig. 8. Clusterig results by the GLA algorithm for DNA-2. come up with much better results tha CL distace i RLS algorithm. It seems that there is o big differece betwee DSC distace ad L 2 - distace whe they are employed i RLS. The differece, however, ca be sigificat whe the correct umber is to be determied i the stepwise search. The result with the GLA is quite differet from that of the RLS, maily because the variace of the results is much greater. The CL distace still performs worse tha the L 2 -distace, but the DSC distace is ow clearly better tha the L 2 -distace almost with respect to every umber of clusters. It is expected that the correct result would be reached more reliably usig the DSC distace. The drawback of the DSC distace is takes much more time to compute tha the L 2 -distace. 5. Coclusios We proposed a ew vector-to-cluster distace i the classificatio problems of biary vectors by miimizig SC. The distace fuctio was applied both i the GLA ad RLS algorithms. Experimets show that RLS by DSC distace gives the best clusterig performace i miimizig SC amog all the variats cosidered, ad the highest classificatio rate. I most cases, the modified CL distace performs eve worse tha the L 2 -distace. It is somehow difficult for the stepwise GLA to deliver satisfactory results i solvig the correct umber of clusters, eve by usig the DSC distace. The L 2 -distace is moderately effective to classify simple data. Amog the three distaces, the DSC distace is the most precise to miimize stochastic complexity. Our approach by usig DSC distace is geeral i its ature as the same desig paradigm ca be applied with ay other cost fuctio too. Table 1 The classificatio results of the RLS ad GLA algorithms with the L 2 -distace, CL-distace ad DSC-distace RLS algorithm GLA algorithm Real classificatio L 2 CL DSC L 2 CL DSC SC Number of a 6 a clusters Classificatio rate 96.98% 92.83% 98.87% 74.34% 87.55% 91.32% 100% a The classificatio rates are calculated from the clusterig of 12 clusters eve though smaller SC-value was foud with 6 clusters.
9 P. Fr ati et al. / Patter Recogitio Letters 24 (2003) Refereces Fr ati, P., Kivij arvi, J., Radomized local search algorithm for the clusterig problem. Patter Aalysis ad Applicatios 3 (4), Fr ati, P., Gylleberg, H.G., Gylleberg, M., Kivij arvi, J., Koski, T., Lud, T., Nevalaie, O., Miimizatio stochastic complexity usig GLA ad local search with applicatios to classificatios of bacteria. Biosystems 57 (1), Gylleberg, M., Koski, T., Probabilistic Models for Bacterial Taxoomy, TUCS Techical Report, No. 325, Turku Ceter for Computer Sciece, Filad, February Gylleberg, M., Koski, M., Verlaa, M., Clusterig ad quatizatio of biary vectors with stochastic complexity. Proc. IEEE Iterat. Symposium o Iformatio Theory, Trodheim, Germay, Gylleberg, M., Koski, T., Verlaa, M., Classificatio of biary vectors by stochastic complexity. Joural of Multivariate Aalysis 63, Kotkae, P., Myllym aki, P., Silader, T., Tirri, H., O the accuracy of stochastic complexity approximatios. I: Gammerma, A. (Ed.), Causal Models ad Itelliget Data Maagemet, Chapter 9. Spriger-Verlag. Lide, Y., Buzo, A., Gray, R., A algorithm for vector quatizer desig. IEEE Tras. Commuicatios 28 (1), Rissae, J., Stochastic Complexity ad the MDL Priciple. Ecoometric Reviews 6, Rissae, J., Stochastic complexity. Joural of Statistics Society 4 (10), Rissae, J., Fisher iformatio ad stochastic complexity. IEEE Tras. o Iformatio Theory 42 (1),
Image Segmentation EEE 508
Image Segmetatio Objective: to determie (etract) object boudaries. It is a process of partitioig a image ito distict regios by groupig together eighborig piels based o some predefied similarity criterio.
More information3D Model Retrieval Method Based on Sample Prediction
20 Iteratioal Coferece o Computer Commuicatio ad Maagemet Proc.of CSIT vol.5 (20) (20) IACSIT Press, Sigapore 3D Model Retrieval Method Based o Sample Predictio Qigche Zhag, Ya Tag* School of Computer
More informationOnes Assignment Method for Solving Traveling Salesman Problem
Joural of mathematics ad computer sciece 0 (0), 58-65 Oes Assigmet Method for Solvig Travelig Salesma Problem Hadi Basirzadeh Departmet of Mathematics, Shahid Chamra Uiversity, Ahvaz, Ira Article history:
More informationAn Efficient Algorithm for Graph Bisection of Triangularizations
A Efficiet Algorithm for Graph Bisectio of Triagularizatios Gerold Jäger Departmet of Computer Sciece Washigto Uiversity Campus Box 1045 Oe Brookigs Drive St. Louis, Missouri 63130-4899, USA jaegerg@cse.wustl.edu
More informationLecture 18. Optimization in n dimensions
Lecture 8 Optimizatio i dimesios Itroductio We ow cosider the problem of miimizig a sigle scalar fuctio of variables, f x, where x=[ x, x,, x ]T. The D case ca be visualized as fidig the lowest poit of
More informationAn Efficient Algorithm for Graph Bisection of Triangularizations
Applied Mathematical Scieces, Vol. 1, 2007, o. 25, 1203-1215 A Efficiet Algorithm for Graph Bisectio of Triagularizatios Gerold Jäger Departmet of Computer Sciece Washigto Uiversity Campus Box 1045, Oe
More informationPattern Recognition Systems Lab 1 Least Mean Squares
Patter Recogitio Systems Lab 1 Least Mea Squares 1. Objectives This laboratory work itroduces the OpeCV-based framework used throughout the course. I this assigmet a lie is fitted to a set of poits usig
More information. Written in factored form it is easy to see that the roots are 2, 2, i,
CMPS A Itroductio to Programmig Programmig Assigmet 4 I this assigmet you will write a java program that determies the real roots of a polyomial that lie withi a specified rage. Recall that the roots (or
More informationHow do we evaluate algorithms?
F2 Readig referece: chapter 2 + slides Algorithm complexity Big O ad big Ω To calculate ruig time Aalysis of recursive Algorithms Next time: Litterature: slides mostly The first Algorithm desig methods:
More informationThe Closest Line to a Data Set in the Plane. David Gurney Southeastern Louisiana University Hammond, Louisiana
The Closest Lie to a Data Set i the Plae David Gurey Southeaster Louisiaa Uiversity Hammod, Louisiaa ABSTRACT This paper looks at three differet measures of distace betwee a lie ad a data set i the plae:
More informationAn Improved Shuffled Frog-Leaping Algorithm for Knapsack Problem
A Improved Shuffled Frog-Leapig Algorithm for Kapsack Problem Zhoufag Li, Ya Zhou, ad Peg Cheg School of Iformatio Sciece ad Egieerig Hea Uiversity of Techology ZhegZhou, Chia lzhf1978@126.com Abstract.
More informationAdministrative UNSUPERVISED LEARNING. Unsupervised learning. Supervised learning 11/25/13. Final project. No office hours today
Admiistrative Fial project No office hours today UNSUPERVISED LEARNING David Kauchak CS 451 Fall 2013 Supervised learig Usupervised learig label label 1 label 3 model/ predictor label 4 label 5 Supervised
More informationChapter 3 Classification of FFT Processor Algorithms
Chapter Classificatio of FFT Processor Algorithms The computatioal complexity of the Discrete Fourier trasform (DFT) is very high. It requires () 2 complex multiplicatios ad () complex additios [5]. As
More informationA New Morphological 3D Shape Decomposition: Grayscale Interframe Interpolation Method
A ew Morphological 3D Shape Decompositio: Grayscale Iterframe Iterpolatio Method D.. Vizireau Politehica Uiversity Bucharest, Romaia ae@comm.pub.ro R. M. Udrea Politehica Uiversity Bucharest, Romaia mihea@comm.pub.ro
More informationPerformance Plus Software Parameter Definitions
Performace Plus+ Software Parameter Defiitios/ Performace Plus Software Parameter Defiitios Chapma Techical Note-TG-5 paramete.doc ev-0-03 Performace Plus+ Software Parameter Defiitios/2 Backgroud ad Defiitios
More informationThe Magma Database file formats
The Magma Database file formats Adrew Gaylard, Bret Pikey, ad Mart-Mari Breedt Johaesburg, South Africa 15th May 2006 1 Summary Magma is a ope-source object database created by Chris Muller, of Kasas City,
More informationAnalysis Metrics. Intro to Algorithm Analysis. Slides. 12. Alg Analysis. 12. Alg Analysis
Itro to Algorithm Aalysis Aalysis Metrics Slides. Table of Cotets. Aalysis Metrics 3. Exact Aalysis Rules 4. Simple Summatio 5. Summatio Formulas 6. Order of Magitude 7. Big-O otatio 8. Big-O Theorems
More informationA Study on the Performance of Cholesky-Factorization using MPI
A Study o the Performace of Cholesky-Factorizatio usig MPI Ha S. Kim Scott B. Bade Departmet of Computer Sciece ad Egieerig Uiversity of Califoria Sa Diego {hskim, bade}@cs.ucsd.edu Abstract Cholesky-factorizatio
More informationBOOLEAN MATHEMATICS: GENERAL THEORY
CHAPTER 3 BOOLEAN MATHEMATICS: GENERAL THEORY 3.1 ISOMORPHIC PROPERTIES The ame Boolea Arithmetic was chose because it was discovered that literal Boolea Algebra could have a isomorphic umerical aspect.
More informationElementary Educational Computer
Chapter 5 Elemetary Educatioal Computer. Geeral structure of the Elemetary Educatioal Computer (EEC) The EEC coforms to the 5 uits structure defied by vo Neuma's model (.) All uits are preseted i a simplified
More informationEvaluation scheme for Tracking in AMI
A M I C o m m u i c a t i o A U G M E N T E D M U L T I - P A R T Y I N T E R A C T I O N http://www.amiproject.org/ Evaluatio scheme for Trackig i AMI S. Schreiber a D. Gatica-Perez b AMI WP4 Trackig:
More informationNew HSL Distance Based Colour Clustering Algorithm
The 4th Midwest Artificial Itelligece ad Cogitive Scieces Coferece (MAICS 03 pp 85-9 New Albay Idiaa USA April 3-4 03 New HSL Distace Based Colour Clusterig Algorithm Vasile Patrascu Departemet of Iformatics
More informationAlgorithms for Disk Covering Problems with the Most Points
Algorithms for Disk Coverig Problems with the Most Poits Bi Xiao Departmet of Computig Hog Kog Polytechic Uiversity Hug Hom, Kowloo, Hog Kog csbxiao@comp.polyu.edu.hk Qigfeg Zhuge, Yi He, Zili Shao, Edwi
More informationExact Minimum Lower Bound Algorithm for Traveling Salesman Problem
Exact Miimum Lower Boud Algorithm for Travelig Salesma Problem Mohamed Eleiche GeoTiba Systems mohamed.eleiche@gmail.com Abstract The miimum-travel-cost algorithm is a dyamic programmig algorithm to compute
More informationBASED ON ITERATIVE ERROR-CORRECTION
A COHPARISO OF CRYPTAALYTIC PRICIPLES BASED O ITERATIVE ERROR-CORRECTIO Miodrag J. MihaljeviC ad Jova Dj. GoliC Istitute of Applied Mathematics ad Electroics. Belgrade School of Electrical Egieerig. Uiversity
More informationCSC165H1 Worksheet: Tutorial 8 Algorithm analysis (SOLUTIONS)
CSC165H1, Witer 018 Learig Objectives By the ed of this worksheet, you will: Aalyse the ruig time of fuctios cotaiig ested loops. 1. Nested loop variatios. Each of the followig fuctios takes as iput a
More informationAnalysis of Documents Clustering Using Sampled Agglomerative Technique
Aalysis of Documets Clusterig Usig Sampled Agglomerative Techique Omar H. Karam, Ahmed M. Hamad, ad Sheri M. Moussa Abstract I this paper a clusterig algorithm for documets is proposed that adapts a samplig-based
More informationFundamentals of Media Processing. Shin'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dinh Le
Fudametals of Media Processig Shi'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dih Le Today's topics Noparametric Methods Parze Widow k-nearest Neighbor Estimatio Clusterig Techiques k-meas Agglomerative Hierarchical
More informationLecture 1: Introduction and Strassen s Algorithm
5-750: Graduate Algorithms Jauary 7, 08 Lecture : Itroductio ad Strasse s Algorithm Lecturer: Gary Miller Scribe: Robert Parker Itroductio Machie models I this class, we will primarily use the Radom Access
More informationStone Images Retrieval Based on Color Histogram
Stoe Images Retrieval Based o Color Histogram Qiag Zhao, Jie Yag, Jigyi Yag, Hogxig Liu School of Iformatio Egieerig, Wuha Uiversity of Techology Wuha, Chia Abstract Stoe images color features are chose
More informationLecture Notes 6 Introduction to algorithm analysis CSS 501 Data Structures and Object-Oriented Programming
Lecture Notes 6 Itroductio to algorithm aalysis CSS 501 Data Structures ad Object-Orieted Programmig Readig for this lecture: Carrao, Chapter 10 To be covered i this lecture: Itroductio to algorithm aalysis
More informationImprovement of the Orthogonal Code Convolution Capabilities Using FPGA Implementation
Improvemet of the Orthogoal Code Covolutio Capabilities Usig FPGA Implemetatio Naima Kaabouch, Member, IEEE, Apara Dhirde, Member, IEEE, Saleh Faruque, Member, IEEE Departmet of Electrical Egieerig, Uiversity
More informationLecture 5. Counting Sort / Radix Sort
Lecture 5. Coutig Sort / Radix Sort T. H. Corme, C. E. Leiserso ad R. L. Rivest Itroductio to Algorithms, 3rd Editio, MIT Press, 2009 Sugkyukwa Uiversity Hyuseug Choo choo@skku.edu Copyright 2000-2018
More informationCIS 121 Data Structures and Algorithms with Java Fall Big-Oh Notation Tuesday, September 5 (Make-up Friday, September 8)
CIS 11 Data Structures ad Algorithms with Java Fall 017 Big-Oh Notatio Tuesday, September 5 (Make-up Friday, September 8) Learig Goals Review Big-Oh ad lear big/small omega/theta otatios Practice solvig
More informationArea As A Limit & Sigma Notation
Area As A Limit & Sigma Notatio SUGGESTED REFERENCE MATERIAL: As you work through the problems listed below, you should referece Chapter 5.4 of the recommeded textbook (or the equivalet chapter i your
More information6.854J / J Advanced Algorithms Fall 2008
MIT OpeCourseWare http://ocw.mit.edu 6.854J / 18.415J Advaced Algorithms Fall 2008 For iformatio about citig these materials or our Terms of Use, visit: http://ocw.mit.edu/terms. 18.415/6.854 Advaced Algorithms
More information9.1. Sequences and Series. Sequences. What you should learn. Why you should learn it. Definition of Sequence
_9.qxd // : AM Page Chapter 9 Sequeces, Series, ad Probability 9. Sequeces ad Series What you should lear Use sequece otatio to write the terms of sequeces. Use factorial otatio. Use summatio otatio to
More informationA SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON
A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON Roberto Lopez ad Eugeio Oñate Iteratioal Ceter for Numerical Methods i Egieerig (CIMNE) Edificio C1, Gra Capitá s/, 08034 Barceloa, Spai ABSTRACT I this work
More informationThe isoperimetric problem on the hypercube
The isoperimetric problem o the hypercube Prepared by: Steve Butler November 2, 2005 1 The isoperimetric problem We will cosider the -dimesioal hypercube Q Recall that the hypercube Q is a graph whose
More informationJournal of Chemical and Pharmaceutical Research, 2013, 5(12): Research Article
Available olie www.jocpr.com Joural of Chemical ad Pharmaceutical Research, 2013, 5(12):745-749 Research Article ISSN : 0975-7384 CODEN(USA) : JCPRC5 K-meas algorithm i the optimal iitial cetroids based
More informationProbabilistic Fuzzy Time Series Method Based on Artificial Neural Network
America Joural of Itelliget Systems 206, 6(2): 42-47 DOI: 0.5923/j.ajis.2060602.02 Probabilistic Fuzzy Time Series Method Based o Artificial Neural Network Erol Egrioglu,*, Ere Bas, Cagdas Haka Aladag
More informationImproving Template Based Spike Detection
Improvig Template Based Spike Detectio Kirk Smith, Member - IEEE Portlad State Uiversity petra@ee.pdx.edu Abstract Template matchig algorithms like SSE, Covolutio ad Maximum Likelihood are well kow for
More informationBezier curves. Figure 2 shows cubic Bezier curves for various control points. In a Bezier curve, only
Edited: Yeh-Liag Hsu (998--; recommeded: Yeh-Liag Hsu (--9; last updated: Yeh-Liag Hsu (9--7. Note: This is the course material for ME55 Geometric modelig ad computer graphics, Yua Ze Uiversity. art of
More informationCS 683: Advanced Design and Analysis of Algorithms
CS 683: Advaced Desig ad Aalysis of Algorithms Lecture 6, February 1, 2008 Lecturer: Joh Hopcroft Scribes: Shaomei Wu, Etha Feldma February 7, 2008 1 Threshold for k CNF Satisfiability I the previous lecture,
More informationPruning and Summarizing the Discovered Time Series Association Rules from Mechanical Sensor Data Qing YANG1,a,*, Shao-Yu WANG1,b, Ting-Ting ZHANG2,c
Advaces i Egieerig Research (AER), volume 131 3rd Aual Iteratioal Coferece o Electroics, Electrical Egieerig ad Iformatio Sciece (EEEIS 2017) Pruig ad Summarizig the Discovered Time Series Associatio Rules
More informationAccuracy Improvement in Camera Calibration
Accuracy Improvemet i Camera Calibratio FaJie L Qi Zag ad Reihard Klette CITR, Computer Sciece Departmet The Uiversity of Aucklad Tamaki Campus, Aucklad, New Zealad fli006, qza001@ec.aucklad.ac.z r.klette@aucklad.ac.z
More informationVALIDATING DIRECTIONAL EDGE-BASED IMAGE FEATURE REPRESENTATIONS IN FACE RECOGNITION BY SPATIAL CORRELATION-BASED CLUSTERING
VALIDATING DIRECTIONAL EDGE-BASED IMAGE FEATURE REPRESENTATIONS IN FACE RECOGNITION BY SPATIAL CORRELATION-BASED CLUSTERING Yasufumi Suzuki ad Tadashi Shibata Departmet of Frotier Iformatics, School of
More informationPerformance Comparisons of PSO based Clustering
Performace Comparisos of PSO based Clusterig Suresh Chadra Satapathy, 2 Guaidhi Pradha, 3 Sabyasachi Pattai, 4 JVR Murthy, 5 PVGD Prasad Reddy Ail Neeruoda Istitute of Techology ad Scieces, Sagivalas,Vishaapatam
More informationIntroduction. Nature-Inspired Computing. Terminology. Problem Types. Constraint Satisfaction Problems - CSP. Free Optimization Problem - FOP
Nature-Ispired Computig Hadlig Costraits Dr. Şima Uyar September 2006 Itroductio may practical problems are costraied ot all combiatios of variable values represet valid solutios feasible solutios ifeasible
More informationPython Programming: An Introduction to Computer Science
Pytho Programmig: A Itroductio to Computer Sciece Chapter 6 Defiig Fuctios Pytho Programmig, 2/e 1 Objectives To uderstad why programmers divide programs up ito sets of cooperatig fuctios. To be able to
More informationProtected points in ordered trees
Applied Mathematics Letters 008 56 50 www.elsevier.com/locate/aml Protected poits i ordered trees Gi-Sag Cheo a, Louis W. Shapiro b, a Departmet of Mathematics, Sugkyukwa Uiversity, Suwo 440-746, Republic
More informationFuzzy Rule Selection by Data Mining Criteria and Genetic Algorithms
Fuzzy Rule Selectio by Data Miig Criteria ad Geetic Algorithms Hisao Ishibuchi Dept. of Idustrial Egieerig Osaka Prefecture Uiversity 1-1 Gakue-cho, Sakai, Osaka 599-8531, JAPAN E-mail: hisaoi@ie.osakafu-u.ac.jp
More informationNew Fuzzy Color Clustering Algorithm Based on hsl Similarity
IFSA-EUSFLAT 009 New Fuzzy Color Clusterig Algorithm Based o hsl Similarity Vasile Ptracu Departmet of Iformatics Techology Tarom Compay Bucharest Romaia Email: patrascu.v@gmail.com Abstract I this paper
More information1.2 Binomial Coefficients and Subsets
1.2. BINOMIAL COEFFICIENTS AND SUBSETS 13 1.2 Biomial Coefficiets ad Subsets 1.2-1 The loop below is part of a program to determie the umber of triagles formed by poits i the plae. for i =1 to for j =
More informationIMP: Superposer Integrated Morphometrics Package Superposition Tool
IMP: Superposer Itegrated Morphometrics Package Superpositio Tool Programmig by: David Lieber ( 03) Caisius College 200 Mai St. Buffalo, NY 4208 Cocept by: H. David Sheets, Dept. of Physics, Caisius College
More informationCSE 417: Algorithms and Computational Complexity
Time CSE 47: Algorithms ad Computatioal Readig assigmet Read Chapter of The ALGORITHM Desig Maual Aalysis & Sortig Autum 00 Paul Beame aalysis Problem size Worst-case complexity: max # steps algorithm
More informationcondition w i B i S maximum u i
ecture 10 Dyamic Programmig 10.1 Kapsack Problem November 1, 2004 ecturer: Kamal Jai Notes: Tobias Holgers We are give a set of items U = {a 1, a 2,..., a }. Each item has a weight w i Z + ad a utility
More information15 UNSUPERVISED LEARNING
15 UNSUPERVISED LEARNING [My father] advised me to sit every few moths i my readig chair for a etire eveig, close my eyes ad try to thik of ew problems to solve. I took his advice very seriously ad have
More informationCounting the Number of Minimum Roman Dominating Functions of a Graph
Coutig the Number of Miimum Roma Domiatig Fuctios of a Graph SHI ZHENG ad KOH KHEE MENG, Natioal Uiversity of Sigapore We provide two algorithms coutig the umber of miimum Roma domiatig fuctios of a graph
More informationComputers and Scientific Thinking
Computers ad Scietific Thikig David Reed, Creighto Uiversity Chapter 15 JavaScript Strigs 1 Strigs as Objects so far, your iteractive Web pages have maipulated strigs i simple ways use text box to iput
More informationCSCI 5090/7090- Machine Learning. Spring Mehdi Allahyari Georgia Southern University
CSCI 5090/7090- Machie Learig Sprig 018 Mehdi Allahyari Georgia Souther Uiversity Clusterig (slides borrowed from Tom Mitchell, Maria Floria Balca, Ali Borji, Ke Che) 1 Clusterig, Iformal Goals Goal: Automatically
More informationChapter 10. Defining Classes. Copyright 2015 Pearson Education, Ltd.. All rights reserved.
Chapter 10 Defiig Classes Copyright 2015 Pearso Educatio, Ltd.. All rights reserved. Overview 10.1 Structures 10.2 Classes 10.3 Abstract Data Types 10.4 Itroductio to Iheritace Copyright 2015 Pearso Educatio,
More informationConvergence results for conditional expectations
Beroulli 11(4), 2005, 737 745 Covergece results for coditioal expectatios IRENE CRIMALDI 1 ad LUCA PRATELLI 2 1 Departmet of Mathematics, Uiversity of Bologa, Piazza di Porta Sa Doato 5, 40126 Bologa,
More informationPseudocode ( 1.1) Analysis of Algorithms. Primitive Operations. Pseudocode Details. Running Time ( 1.1) Estimating performance
Aalysis of Algorithms Iput Algorithm Output A algorithm is a step-by-step procedure for solvig a problem i a fiite amout of time. Pseudocode ( 1.1) High-level descriptio of a algorithm More structured
More informationData Structures Week #9. Sorting
Data Structures Week #9 Sortig Outlie Motivatio Types of Sortig Elemetary (O( 2 )) Sortig Techiques Other (O(*log())) Sortig Techiques 21.Aralık.2010 Boraha Tümer, Ph.D. 2 Sortig 21.Aralık.2010 Boraha
More informationAlpha Individual Solutions MAΘ National Convention 2013
Alpha Idividual Solutios MAΘ Natioal Covetio 0 Aswers:. D. A. C 4. D 5. C 6. B 7. A 8. C 9. D 0. B. B. A. D 4. C 5. A 6. C 7. B 8. A 9. A 0. C. E. B. D 4. C 5. A 6. D 7. B 8. C 9. D 0. B TB. 570 TB. 5
More informationDynamic Programming and Curve Fitting Based Road Boundary Detection
Dyamic Programmig ad Curve Fittig Based Road Boudary Detectio SHYAM PRASAD ADHIKARI, HYONGSUK KIM, Divisio of Electroics ad Iformatio Egieerig Chobuk Natioal Uiversity 664-4 Ga Deokji-Dog Jeoju-City Jeobuk
More informationAPPLICATION NOTE PACE1750AE BUILT-IN FUNCTIONS
APPLICATION NOTE PACE175AE BUILT-IN UNCTIONS About This Note This applicatio brief is iteded to explai ad demostrate the use of the special fuctios that are built ito the PACE175AE processor. These powerful
More informationCubic Polynomial Curves with a Shape Parameter
roceedigs of the th WSEAS Iteratioal Coferece o Robotics Cotrol ad Maufacturig Techology Hagzhou Chia April -8 00 (pp5-70) Cubic olyomial Curves with a Shape arameter MO GUOLIANG ZHAO YANAN Iformatio ad
More informationANN WHICH COVERS MLP AND RBF
ANN WHICH COVERS MLP AND RBF Josef Boští, Jaromír Kual Faculty of Nuclear Scieces ad Physical Egieerig, CTU i Prague Departmet of Software Egieerig Abstract Two basic types of artificial eural etwors Multi
More informationSolution printed. Do not start the test until instructed to do so! CS 2604 Data Structures Midterm Spring, Instructions:
CS 604 Data Structures Midterm Sprig, 00 VIRG INIA POLYTECHNIC INSTITUTE AND STATE U T PROSI M UNI VERSI TY Istructios: Prit your ame i the space provided below. This examiatio is closed book ad closed
More informationCh 9.3 Geometric Sequences and Series Lessons
Ch 9.3 Geometric Sequeces ad Series Lessos SKILLS OBJECTIVES Recogize a geometric sequece. Fid the geeral, th term of a geometric sequece. Evaluate a fiite geometric series. Evaluate a ifiite geometric
More informationLecturers: Sanjam Garg and Prasad Raghavendra Feb 21, Midterm 1 Solutions
U.C. Berkeley CS170 : Algorithms Midterm 1 Solutios Lecturers: Sajam Garg ad Prasad Raghavedra Feb 1, 017 Midterm 1 Solutios 1. (4 poits) For the directed graph below, fid all the strogly coected compoets
More informationCOSC 1P03. Ch 7 Recursion. Introduction to Data Structures 8.1
COSC 1P03 Ch 7 Recursio Itroductio to Data Structures 8.1 COSC 1P03 Recursio Recursio I Mathematics factorial Fiboacci umbers defie ifiite set with fiite defiitio I Computer Sciece sytax rules fiite defiitio,
More informationChapter 5. Functions for All Subtasks. Copyright 2015 Pearson Education, Ltd.. All rights reserved.
Chapter 5 Fuctios for All Subtasks Copyright 2015 Pearso Educatio, Ltd.. All rights reserved. Overview 5.1 void Fuctios 5.2 Call-By-Referece Parameters 5.3 Usig Procedural Abstractio 5.4 Testig ad Debuggig
More informationarxiv: v2 [cs.ds] 24 Mar 2018
Similar Elemets ad Metric Labelig o Complete Graphs arxiv:1803.08037v [cs.ds] 4 Mar 018 Pedro F. Felzeszwalb Brow Uiversity Providece, RI, USA pff@brow.edu March 8, 018 We cosider a problem that ivolves
More informationChapter 11. Friends, Overloaded Operators, and Arrays in Classes. Copyright 2014 Pearson Addison-Wesley. All rights reserved.
Chapter 11 Frieds, Overloaded Operators, ad Arrays i Classes Copyright 2014 Pearso Addiso-Wesley. All rights reserved. Overview 11.1 Fried Fuctios 11.2 Overloadig Operators 11.3 Arrays ad Classes 11.4
More informationBig-O Analysis. Asymptotics
Big-O Aalysis 1 Defiitio: Suppose that f() ad g() are oegative fuctios of. The we say that f() is O(g()) provided that there are costats C > 0 ad N > 0 such that for all > N, f() Cg(). Big-O expresses
More informationn n B. How many subsets of C are there of cardinality n. We are selecting elements for such a
4. [10] Usig a combiatorial argumet, prove that for 1: = 0 = Let A ad B be disjoit sets of cardiality each ad C = A B. How may subsets of C are there of cardiality. We are selectig elemets for such a subset
More informationChapter 1. Introduction to Computers and C++ Programming. Copyright 2015 Pearson Education, Ltd.. All rights reserved.
Chapter 1 Itroductio to Computers ad C++ Programmig Copyright 2015 Pearso Educatio, Ltd.. All rights reserved. Overview 1.1 Computer Systems 1.2 Programmig ad Problem Solvig 1.3 Itroductio to C++ 1.4 Testig
More informationProject 2.5 Improved Euler Implementation
Project 2.5 Improved Euler Implemetatio Figure 2.5.10 i the text lists TI-85 ad BASIC programs implemetig the improved Euler method to approximate the solutio of the iitial value problem dy dx = x+ y,
More informationDATA MINING II - 1DL460
DATA MINING II - 1DL460 Sprig 2017 A secod course i data miig http://www.it.uu.se/edu/course/homepage/ifoutv2/vt17/ Kjell Orsbor Uppsala Database Laboratory Departmet of Iformatio Techology, Uppsala Uiversity,
More informationWhat are we going to learn? CSC Data Structures Analysis of Algorithms. Overview. Algorithm, and Inputs
What are we goig to lear? CSC316-003 Data Structures Aalysis of Algorithms Computer Sciece North Carolia State Uiversity Need to say that some algorithms are better tha others Criteria for evaluatio Structure
More informationImproving Information Retrieval System Security via an Optimal Maximal Coding Scheme
Improvig Iformatio Retrieval System Security via a Optimal Maximal Codig Scheme Dogyag Log Departmet of Computer Sciece, City Uiversity of Hog Kog, 8 Tat Chee Aveue Kowloo, Hog Kog SAR, PRC dylog@cs.cityu.edu.hk
More informationChapter 9. Pointers and Dynamic Arrays. Copyright 2015 Pearson Education, Ltd.. All rights reserved.
Chapter 9 Poiters ad Dyamic Arrays Copyright 2015 Pearso Educatio, Ltd.. All rights reserved. Overview 9.1 Poiters 9.2 Dyamic Arrays Copyright 2015 Pearso Educatio, Ltd.. All rights reserved. Slide 9-3
More informationData Structures and Algorithms. Analysis of Algorithms
Data Structures ad Algorithms Aalysis of Algorithms Outlie Ruig time Pseudo-code Big-oh otatio Big-theta otatio Big-omega otatio Asymptotic algorithm aalysis Aalysis of Algorithms Iput Algorithm Output
More informationMATHEMATICAL METHODS OF ANALYSIS AND EXPERIMENTAL DATA PROCESSING (Or Methods of Curve Fitting)
MATHEMATICAL METHODS OF ANALYSIS AND EXPERIMENTAL DATA PROCESSING (Or Methods of Curve Fittig) I this chapter, we will eamie some methods of aalysis ad data processig; data obtaied as a result of a give
More informationEuclidean Distance Based Feature Selection for Fault Detection Prediction Model in Semiconductor Manufacturing Process
Vol.133 (Iformatio Techology ad Computer Sciece 016), pp.85-89 http://dx.doi.org/10.1457/astl.016. Euclidea Distace Based Feature Selectio for Fault Detectio Predictio Model i Semicoductor Maufacturig
More informationAN OPTIMIZATION NETWORK FOR MATRIX INVERSION
397 AN OPTIMIZATION NETWORK FOR MATRIX INVERSION Ju-Seog Jag, S~ Youg Lee, ad Sag-Yug Shi Korea Advaced Istitute of Sciece ad Techology, P.O. Box 150, Cheogryag, Seoul, Korea ABSTRACT Iverse matrix calculatio
More informationA Parallel DFA Minimization Algorithm
A Parallel DFA Miimizatio Algorithm Ambuj Tewari, Utkarsh Srivastava, ad P. Gupta Departmet of Computer Sciece & Egieerig Idia Istitute of Techology Kapur Kapur 208 016,INDIA pg@iitk.ac.i Abstract. I this
More informationEE 459/500 HDL Based Digital Design with Programmable Logic. Lecture 13 Control and Sequencing: Hardwired and Microprogrammed Control
EE 459/500 HDL Based Digital Desig with Programmable Logic Lecture 13 Cotrol ad Sequecig: Hardwired ad Microprogrammed Cotrol Refereces: Chapter s 4,5 from textbook Chapter 7 of M.M. Mao ad C.R. Kime,
More informationNeuro Fuzzy Model for Human Face Expression Recognition
IOSR Joural of Computer Egieerig (IOSRJCE) ISSN : 2278-0661 Volume 1, Issue 2 (May-Jue 2012), PP 01-06 Neuro Fuzzy Model for Huma Face Expressio Recogitio Mr. Mayur S. Burage 1, Prof. S. V. Dhopte 2 1
More informationA Novel Feature Extraction Algorithm for Haar Local Binary Pattern Texture Based on Human Vision System
A Novel Feature Extractio Algorithm for Haar Local Biary Patter Texture Based o Huma Visio System Liu Tao 1,* 1 Departmet of Electroic Egieerig Shaaxi Eergy Istitute Xiayag, Shaaxi, Chia Abstract The locality
More informationThe number n of subintervals times the length h of subintervals gives length of interval (b-a).
Simulator with MadMath Kit: Riema Sums (Teacher s pages) I your kit: 1. GeoGebra file: Ready-to-use projector sized simulator: RiemaSumMM.ggb 2. RiemaSumMM.pdf (this file) ad RiemaSumMMEd.pdf (educator's
More informationNumerical Methods Lecture 6 - Curve Fitting Techniques
Numerical Methods Lecture 6 - Curve Fittig Techiques Topics motivatio iterpolatio liear regressio higher order polyomial form expoetial form Curve fittig - motivatio For root fidig, we used a give fuctio
More informationComputer Science Foundation Exam. August 12, Computer Science. Section 1A. No Calculators! KEY. Solutions and Grading Criteria.
Computer Sciece Foudatio Exam August, 005 Computer Sciece Sectio A No Calculators! Name: SSN: KEY Solutios ad Gradig Criteria Score: 50 I this sectio of the exam, there are four (4) problems. You must
More informationRecursive Procedures. How can you model the relationship between consecutive terms of a sequence?
6. Recursive Procedures I Sectio 6.1, you used fuctio otatio to write a explicit formula to determie the value of ay term i a Sometimes it is easier to calculate oe term i a sequece usig the previous terms.
More information2D Isogeometric Shape Optimization considering both control point positions and weights as design variables
1 th World Cogress o tructural ad Multidiscipliary Optimizatio May 19-24, 213, Orlado, Florida, UA 2D Isogeometric hape Optimizatio cosiderig both cotrol poit positios ad weights as desig variables Yeo-Ul
More informationOn Infinite Groups that are Isomorphic to its Proper Infinite Subgroup. Jaymar Talledo Balihon. Abstract
O Ifiite Groups that are Isomorphic to its Proper Ifiite Subgroup Jaymar Talledo Baliho Abstract Two groups are isomorphic if there exists a isomorphism betwee them Lagrage Theorem states that the order
More informationLecture 6. Lecturer: Ronitt Rubinfeld Scribes: Chen Ziv, Eliav Buchnik, Ophir Arie, Jonathan Gradstein
068.670 Subliear Time Algorithms November, 0 Lecture 6 Lecturer: Roitt Rubifeld Scribes: Che Ziv, Eliav Buchik, Ophir Arie, Joatha Gradstei Lesso overview. Usig the oracle reductio framework for approximatig
More information