A Support-Based Algorithm for the Bi-Objective Pareto Constraint

Size: px
Start display at page:

Download "A Support-Based Algorithm for the Bi-Objective Pareto Constraint"

Transcription

1 A Support-Based Algorithm for the Bi-Ojetive Pareto Constraint Renaud Hartert and Pierre Shaus UCLouvain, ICTEAM, Plae Sainte Bare 2, 1348 Louvain-la-Neuve, Belgium {renaud.hartert, Astrat Bi-Ojetive Cominatorial Optimization prolems are uiquitous in real-world appliations and designing approahes to solve them effiiently is an important researh area of Artifiial Intelligene. In Constraint Programming, the reently introdued i-ojetive Pareto onstraint allows one to solve i-ojetive ominatorial optimization prolems exatly. Using this onstraint, every non-dominated solution is olleted in a single tree-searh while pruning su-trees that annot lead to a non-dominated solution. This paper introdues a simpler and more effiient filtering algorithm for the i-ojetive Pareto onstraint. The effiieny of this algorithm is experimentally onfirmed on lassial i-ojetive enhmarks. Bi-Ojetive Cominatorial Optimization (BOCO) aims at optimizing two ojetive funtions simultaneously. Sine these ojetive funtions are often onfliting, there is usually no perfet solution that is optimal for oth ojetives at the same time. In this ontext, deision makers are looking for all the est ompromises etween the ojetives to hoose a posteriori the solution that est fits their needs. Hene, the notion of optimal solution is replaed y the notion of effiieny and we are searhing for the set of all the effiient solutions (usually alled effiient set or Pareto frontier) instead of one single solution (Ehrgott 2005). During the past years, many approahes were developed to takle BOCO prolems exatly. However, many of them were developed in the ontext of Mathematial Programming (Mavrotas 2007; Ralphs, Saltzman, and Wieek 2006) and only a few an e applied effiiently in Constraint Programming. Among these approahes, the ɛ-onstraint is proaly the most widely used (Haimes, Lasdon, and Wismer 1971; Le Pape et al. 1994; Van Wassenhove and Gelders 1980). The idea is to deompose the original prolem into a sequene of suprolems to optimize with regard to the first ojetive funtion. At eah iteration, a new suprolem is generated y onstraining the seond ojetive to take a etter value than its value in the optimal solution of the previously solved suprolem. Notie that the numer of single- This work was (partially) supported y the ARC grant 13/ from Communauté française de Belgique. Copyright 2014, Assoiation for the Advanement of Artifiial Intelligene ( All rights reserved. ojetive prolems to solve is linear in the effiient set s ardinality (Haimes, Lasdon, and Wismer 1971). The i-ojetive Pareto onstraint is an alternative (and more effiient (Gavanelli 2002)) approah to ompute the effiient set exatly. The idea ehind this onstraint is to uild an approximation of the effiient set inrementally during the searh and to use this approximation to detet and to prune su-trees that an only lead to solutions that are less effiient than the ones already ontained in the approximation. Eventually, the approximation eomes the effiient set and its optimality is proven when the searh is ompleted. The algorithm of the Pareto onstraint relies on two operations. The first operation is used to update the approximation (y inserting new solutions in it) while the seond onsists to use this approximation to redue the searh spae of the prolem. In this work, we show how to use speifi BOCO properties to improve the effiieny of the iojetive Pareto onstraint. Preisely, we show that oth operations (update and filtering) an enefit from eah other in an iterative way to uild a simpler and more effiient algorithm for the onstraint. This doument is strutured as follows. First, we riefly introdue onstraint programming and its main onepts. Then, we formalize multi-ojetive ominatorial optimization in the ontext of onstraint programming and present some important definitions. The third setion is dediated to the Pareto onstraint in its general multi-ojetive form. Our support-ased algorithm is presented in the fourth setion. Setion five diretly follows with our experiments and results on two lassial enhmarks i.e. the i-ojetive knapsak prolem and the i-ojetive travelling salesman prolem. Finally, the last setion offers some onlusions and perspetives. Constraint Programming Bakground Constraint Programming is a powerful paradigm to solve onstraint satisfation prolems and ominatorial optimization prolems. A onstraint programming prolem is usually defined y a set of variales with their respetive domain (i.e. the set of values that an e assigned to a variale), and a set of onstraints on these variales. The ojetive is to find an assignment of the variales that respets all the onstraints of the prolem. The onstraint programming proess interleaves a tree-searh exploration (ommon in ar-

2 tifiial intelligene) with an inferene proedure (also alled propagation or filtering) to remove values that annot appear in a solution. The inferene part prunes the ranhes of the searh-tree and thus redues the searh spae to explore. In an optimization ontext, an ojetive funtion an also e onsidered. Usually, this is formulated y the addition of an ojetive variale having as domain the set of possile values that an e taken y the ojetive funtion. Eah time a solution is found, a onstraint is added to the prolem dynamially to fore the ojetive variale to e assigned to a etter value. 1 The optimality is proven when the searh is exhausted, the last solution eing the optimal one. Multi-Ojetive Cominatorial Optimization A Multi-Ojetive Cominatorial Optimization (MOCO) prolem is a quadruple P = X, D, C, F where X = {x 1,..., x n } is a set of variales, D = {D 1,..., D n } is a set of domains, C is a set of onstraints on the variales in X, and F = {f 1,..., f m } is a set of ojetive funtions to minimize simultaneously. Eah ojetive funtion f i (X) assoiates a disrete ost to the assignment of the variales in X. In the following, we assume that m ojetive variales oj 1,..., oj m have een added to the set of variales X and onstrained to e equal to the value taken y their orresponding ojetive funtion i.e. oj i = f i (X). In partiular, the minimal and maximal values of the domain of an ojetive variale oj i are the ounds of this ojetive with regard to a given partial assignment of the variales in X. A solution of a MOCO prolem P is a omplete assignment of the variales in X that satisfies all the onstraints in C. In this work, we represent a solution sol y its assoiated ojetive vetor (sol 1,..., sol m ) where sol i is the value assigned to the orresponding ojetive variales oj i. As it is not likely that a solution is simultaneously optimal for all ojetives at the same time, we are interested in a way to define a partial order on the ojetive vetor of the solutions. Among these orderings, the weak Pareto dominane is a popular hoie (see Figure 1). It is defined as follows. Definition 1 (Weak Pareto dominane). Consider sol = (sol 1,..., sol m ) and sol = (sol 1,..., sol m), two solutions of a MOCO prolem P. We say that sol dominates sol, denoted sol sol, if and only if: i {1,..., m} : sol i sol i. (1) Let sols(p) denote the set of all the solutions of a MOCO prolem P. A solution sol of P is effiient if and only if no solution sol in sols(p) dominates sol: sol sols(p) : sol sol sol sol. (2) In other words, a solution is effiient if it is impossile to improve the value of one ojetive without degrading the value of at least one other ojetive (i.e. an effiient solution is an optimal ompromise etween the ojetives). Solving a MOCO prolem usually means finding the set of all the effiient solutions that is generally alled the effiient set or Pareto front (the effiient set of an aritrary solution spae is depited in Figure 2). Unfortunately, the size 1 Notie that the searh does not stop ut ontinues after the addition of the onstraint to the prolem. of the effiient set often grows exponentially with the size of the prolem to solve (Ehrgott 2005). Hene, in pratie, only an approximation of the effiient set an e found in a reasonale amount of time and of memory. We all suh an approximation of the effiient set an arhive. Definition 2 (Arhive). An arhive A is a domination-free set of solutions i.e. a set of solutions suh that no solution in A dominates another solution in A: sol A, sol A : sol sol sol sol. (3) Clearly, the effiient set is also an arhive. a f e Figure 1: Solution e dominates the solutions,, and d while eing dominated y solution f. Solutions a and g do not dominate solution e and are not dominated y solution e. g d The Pareto Constraint Figure 2: An aritrary iojetive solution spae. Effiient solutions are olored in lak. The effiient set is the set of all the effiient solutions. The Pareto onstraint (Shaus and Hartert 2013) is a gloal onstraint defined over the m ojetive variales of a MOCO prolem P and an arhive A: Pareto(oj 1,..., oj m, A). (4) The aim of the Pareto onstraint is to use the arhive A to prune solutions that are dominated y at least one solution ontained in the arhive A. In other words, the Pareto onstraint ensures that a newly disovered solution is not dominated y any solution in the arhive. This definition is formalized as follows: m oj i < sol i (5) Filtering sol A i=1 The filtering rule of the Pareto onstraint was originally proposed in (Gavanelli 2002). The idea is to use artifiial ojetive vetors, alled dominated points, to adjust the upper ounds of the ojetive variales in order to prevent the disovery of dominated solutions (5). Definition 3 (Dominated point). Let oji min and oji max denote the lower and upper ounds of the ojetive variale oj i, the dominated point DP i is defined as follows:

3 where DP i = (DP i 1,..., DP i m) { oj DPj i max = j ojj min if j = i otherwise In other words, the dominated point DP i an e seen as an artifiial solution for whih eah ojetive variale is assigned to its est possile value exept for oj i whih is assigned to its worst possile value. Proposition 1. Let sol e a solution in the arhive A suh that sol dominates DP i. Then, the value of the ojetive variale oj i has to e smaller than the value sol i : (6) sol A, sol DP i oj i < sol i. (7) Proof. Consider a solution sol in the arhive A suh that sol DP i. Aording to (6), we know that sol j ojj min ( j {1,..., m}, j i). Hene, eah newly disovered solution sol new (i.e. a solution ontained in the Cartesian produt of the ojetive variales) suh that sol i soli new is dominated y sol and an thus e safely removed. Figure 3 illustrates the domain of the ojetive variales efore (left part) and after (right part) applying the filtering rule of Proposition 1. DP 1 DP 1 Figure 3: Filtering of the Pareto onstraint. Blak solutions orrespond to the solutions ontained in the arhive A. Grey areas represent the parts of the ojetive spae that are dominated y at least one solution in A. The dominated points DP 1 and are represented y the white solutions. The hashed area orresponds to the redution of the ojetive variales domains after applying the filtering rule of Proposition 1. If a dominated point DP i is dominated y several solutions at the same time, the filtering rule of Proposition 1 has to e applied until no solution in the arhive dominates DP i. Clearly, the order in whih the dominating solutions are seleted affets the numer of alls of the filtering rule. This situation is illustrated in Figure 4 where the worst possile seletion order is a,, and. From this oservation, it appears that the Pareto onstraint an reah its fixed point in one step if it is ale to aess diretly the solution that dominates DP i with the lowest value in oj i. Suh a solution is alled the tightest solution of oj i. Definition 4 (Tightest solution). The tightest solution of oj i is the solution in the arhive A that dominates DP i and that has the lowest value for oj i : argmin sol A {sol i sol DP i }. (8) If DP i is not dominated y at least one solution in A, then, oj i has no tightest solution. Tightest solutions have een used in (Shaus and Hartert 2013) to propose the following idempotent filtering rule: sol A, sol DP i oj i < T i i (9) where T i is the tightest solution of oj i. a Figure 4: Worst possile seletion order. Solution is the tightest solution of oj 2. Multi-Ojetive Branh-and-Bound The Pareto onstraint an e used to find the exat effiient set of any MOCO prolem. The idea is to improve the quality of the arhive A used y the onstraint dynamially during the searh. Initially, the arhive is the empty set. Then, eah time a new solution is disovered, it is inserted into the arhive and the searh ontinues. Aording to (5), every newly disovered solution is guaranteed to e dominated y none of the solutions in the arhive. Hene, a newly disovered solution inserted in the arhive improves its quality and onsequently strengthens the filtering. Two situations are possile when inserting a new solution in the arhive: 1. The new solution does not dominate any solution in the arhive and nothing has to e done after the insertion (see Figure 5 left); 2. The new solution dominates one or several solutions in the arhive. In this ase, these dominated solutions have to e removed from the arhive to ensure that the arhive remains domination-free (see Figure 5 right). In oth ases, adding the new solution in the arhive stritly inreases the size of the suspae of the ojetive spae that is dominated y the arhive. 2 Eventually, the arhive eomes the effiient set and its optimality is proven when the searh is exhausted. The use of the Pareto onstraint with a dynamially improving arhive an e seen as a multiojetive ranh-and-ound that generalizes the lassial ranh-and-ound algorithms used in onstraint programming (Gavanelli 2002). 3 2 The size of the suspae that is dominated y an arhive is a ommon indiator, known as the Hypervolume, used to measure the quality of an arhive (Zitzler et al. 2003). 3 In single ojetive optimization, the arhive is either the empty set or a singleton ontaining the est-so-far solution. The only dominated point DP 1 orresponds to the upper ound of the ojetive variale to minimize.

4 B C A sol new new sol F D G A E B C D E F G Figure 5: Insertion of a new solution in an arhive. Hashed areas orrespond to the additional parts of the ojetive spae that are dominated after the insertion of the new solution in the arhive. Implementation and data struture The multi-ojetive ranh-and-ound algorithm ased on the Pareto onstraint relies on two distint operations: Aess the tightest solution of eah ojetive to adjust the upper ound of the ojetive variales (9); Insert a newly disovered solution in the arhive and remove potentially dominated solutions from the arhive. Clearly, the effiieny of these operations is impated y the underlying data struture used to implement the inner mehanisms of the onstraint and to store its arhive. In (Gavanelli 2002), the author suggests the use of point quad trees (simply quad-trees in the sequel) to implement the arhive. A quad-tree (Finkel and Bentley 1974; Haeniht 1983; Samet 2006) is a data struture that generalizes inary searh trees to store m-dimensional vetors. Any node of a quad-tree divides its suspae into 2 m disjoint suspaes. More preisely, the root of the tree divides the spae into 2 m suspaes originating at the root. The 2 m hildren of the root (exatly one for eah suspae of the root) divides their suspaes into 2 m suspaes and so on (see Figure 6). As for inary searh trees, quad-trees are very sensitive to the order in whih solutions are inserted. In the est ase, when the tree is well-alaned, a quad-tree ensures a logarithmi omplexity for the aess operation. In the worst ase however, the quad-tree is strutured as a linear list and the aess operation needs to traverse the entire data struture to find the tightest solutions. The removal operations are also a weakness of this data struture. Indeed, removing a solution in a quad-tree leads to the destrution of the sutree rooted at this solution and often leads to an expensive omputational ost to repair it. Unfortunately, the insert operation may require several expensive removals to maintain the arhive domination-free (i.e. to remove the solutions that are dominated y the inserted solution). Support-Based Algorithms Bi-Ojetive Cominatorial Optimization (BOCO) prolems are MOCO prolems with only two ojetives (m = 2). Usually, BOCO prolems are easier to reason aout and have properties that annot e generalized to MOCO prolems with more than two ojetives. One of these properties Figure 6: Illustration of the spae partition (left) of a idimensional quad-tree (right). The node A is the root of the quad-tree. Nodes B, C, D, and E are the sons of A. Nodes F and G are the sons of D. is the i-ojetive ordering property that is defined as follows. Property 1 (i-ojetive ordering). Let A e an aritrary i-ojetive arhive. We denote A >i (resp. A <i ) the arhive A ordered y dereasing (resp. inreasing) value of oj i. Then, sorting the solutions of A y dereasing order of their value in a first ojetive (i.e. A >1 ) amounts to sorting these solutions y inreasing order of their value in the seond ojetive (i.e. A <2 ) and vie-versa. Proof. Let sol and sol e two solutions in A. Sine A is domination-free, sol does not dominate sol and sol does not dominate sol. Hene, if sol 1 < sol 1 we know that sol 2 > sol 2. Symmetrially, if sol 1 > sol 1 then sol 2 < sol 2. In the remainder of this setion, we introdue an inremental algorithm to implement the i-ojetive Pareto onstraint. The idea ehind this algorithm is to use the iojetive ordering property to store the arhive A in a iordered linked-list i.e. a linked-list suh that eah solution has a pointer to its diret suessor in A >1 and in A >2 (see Figure 7). Preisely, the algorithm relies on speial solutions, alled supports, to maintain the tightest solutions of eah ojetive (Definition 4) inrementally during the exploration of the searh tree. Definition 5 (Support). Let sol e a solution in a iojetive arhive A. We say that sol is the support of oj 1 if and only if: sol = argmin sol A {sol 1 sol 2 < oj min 2 }. (10) The support of oj 2 is defined symmetrially. Figure 8 illustrates the supports of oth ojetive variales oj 1 and oj 2. Proposition 2. Supports are never inluded in the Cartesian produt of the domain of the ojetive variales i.e. supports annot e dominated y newly disovered solutions. Proof. Let sol sup e the support of oj 1 (resp. oj 2 ). By Definition 5, we know that sol sup 2 < oj2 min (resp. sol sup 1 < oj1 min ). Hene, sol sup annot e dominated y a newly disovered solution that is at est (oj1 min, oj2 min ).

5 a Figure 7: Illustration of an aritrary i-ojetive arhive A stored in a i-ordered linked-list. Eah solution has a pointer to its diret suessor in A >1 and in A >2. d e a Figure 8: Solutions and e are respetively the supports of oj 2 and oj 1. Both supports are not inluded in the Cartesian produt of the ojetive variales. Proposition 3. If it exists, the tightest solution of oj i is the support of oj i or its diret suessor in A >i. Proof. Let DP 1 e dominated y a solution sol suh that sol is the tightest solution of oj 1. If sol 2 < oj2 min, y Definition 5 we know that sol is the support of oj 1. Else, sol 2 = oj2 min and sol is the diret suessor of the support of oj 1. The link etween supports and tightest solutions is illustrated in Figure 8 where d is the tightest solution of oj 1 and is the tightest solution of oj 2. Proposition 4. Let sol new = (oj1 min, oj2 min ) e a newly disovered solution. If the arhive is ordered aording to the value of one ojetive (i.e. A >1 or A >2 ), then, sol new dominates all the solutions ontained etween oth supports. Proof. Let S i denote the set of all the suessors of the support of oj i in the ordered arhive A >i. By Definition 5, we know that S i = {sol A sol i oji min }. The intersetion of S 1 and S 2 is thus the set of solutions that is dominated y (oj1 min, oj2 min ): S 1 S 2 = {sol A sol 1 oj min 1 sol 2 oj min 2 }. Sine sol new = (oj1 min, oj2 min ), the solutions ontained in the intersetion of the suessors of oth supports are dominated y sol new. The insertion of a new solution is illustrated in Figure 9. This operation is performed in onstant time y updating the pointers of oth supports. We use Propositions 2, 3, and 4 to design an inremental algorithm for the i-ojetive Pareto onstraint. As mentioned aove, the algorithm maintains the tightest solutions of eah ojetive y adjusting the supports inrementally during the exploration of the searh tree. More preisely, we desrie the algorithm to adjust the upper ound of oj 1 as follows: 4 4 The upper ound of oj 2 is adjusted symmetrially. d e 1. Eah time the lower ound of oj 2 is adjusted, we know that we have to reonsider the support of oj 1. To do so, we iterate on the diret suessors of the old support in A >1 until we reah the new support. Let denote this numer of iterations. Clearly, the sum of the annot exeed the size of A along a ranh of the searh tree. 2. When the new support is found, we aess and use the tightest solution of oj 1 to adjust the upper ound of oj 1 (see Proposition 3 and (9)). 3. When a new solution sol new is disovered, we insert it in A y simply updating the pointers of its supports (see Proposition 4). Assuming a trail-ased CP solver, reversile pointers an e used to maintain the supports. 5 Hene, eah time a aktrak ours, the algorithm is ale to restore its previous supports in amortized onstant time. Oserve that, aording to Proposition 2, we know that these supports are not dominated y a previously disovered solution. s 2 sol new s 1 s 2 sol new Figure 9: Insertion of a new solution in onstant time. Solutions s 1 and s 2 are respetively the support of oj 1 and oj 2. Experiments and Results This setion presents the experimental evaluation of our support-ased algorithm against the quad-tree algorithm ommonly used to implement the i-ojetive Pareto onstraint. Our experiments used lassial instanes of the i-ojetive inary knapsak prolem (Xavier Gandileux 2013) and instanes of the i-ojetive traveling salesman prolem (Paquete and Stützle 2009). All algorithms were implemented in the open-soure OsaR solver (OsaR Team 2012) that runs on the Java Virtual Mahine using a omputer running Ma Os X 10.9 on an Intel i7 2.6 Ghz proessor. First, we ompare the numer of searh nodes explored using oth algorithms on instanes of the i-ojetive inary knapsak prolem within a time limit of 30 seonds. In this experiment, the i-ojetive Pareto onstraint starts with an empty arhive and explores the searh-spae with a random heuristi. The usage of a random heuristi should favor the quad-tree, highly sensitive to the order in whih solutions are inserted. Indeed, using a heuristi dediated to one of the ojetives would lead to an unalaned quad-tree negatively 5 This funtionality an easily e adapted for opy-ased CP Solvers. s 1

6 impating its effiieny. Tale 1 gives the mean and the standard deviation of the numer of nodes explored over 10 runs for eah algorithm. We oserve that our support-ased algorithm is always the fastest despite the fair heuristi that we have hosen. It explores on average 20% more nodes. Tale 1: Numer of searh nodes explored within a time limit of 30 seonds on instanes of the i-ojetive inary knapsak prolem with 500 items. Quad-tree Support-ased instane mean st. dev. mean st. dev. 500A B C D Our seond experiment ompares oth algorithms on instanes of the i-ojetive traveling salesman prolem. In this experiments, the Pareto onstraint starts with an initial arhive that is a good approximation of the exat effiient set. In fat, the effiient set of these instane is urrently unknown and the given set is the union of the approximation of many state-of-the-art algorithm to solve this prolem (Lust and Teghem 2010). Sine the internal struture of the Pareto onstraint is initialized in advane, we onsider oth ases of an unalaned and well-alaned quad-tree. As for the previous experiments, the mean and standard deviation over 10 runs are presented in the Tale 2. Tale 2: Numer of searh nodes explored within a time limit of 30 seonds on instanes of the i-ojetive travelling salesman prolem with 100, 150, 200, 300, 400, and 500 ities. Quad-tree Quad-tree (al.) Support-ased instane mean st. dev. mean st. dev. mean st. dev. KroAB KroAB KroAB KroAB KroAB KroAB Again, the support-ased is the fastest approah. As mentioned aove, we oserve that the quad-tree algorithm is very sensitive to the order in whih solutions are added in its struture. On one hand, a well-alaned quad-tree is ompetitive with the support-ased algorithm (ut still always dominated). On the other hand, a poorly alaned quad-tree sustantially deteriorates the numer of searh nodes explored. Interestingly, the alaned quad-tree and the support-ased algorithms were ale to add 3 solutions in the approximation of the KroAB300 effiient set and 4 solutions in the approximation of the KroAB500 effiient set. This experiment illustrates the flexiility of the Pareto onstraint and one of its possile uses to improve an existing approximation of the effiient set or to prove its optimality. Conlusion This paper introdued a support-ased algorithm to implement the i-ojetive Pareto onstraint. This inremental filtering algorithm relies on the i-ojetive ordering property of BOCO prolems. Experiments demonstrate that the support-ased algorithm is more effiient than the lassial algorithm used to implement the Pareto onstraint. Also, this algorithm is simpler to implement than the quad-tree ased propagator sine it only relies on a linked list and two reversile pointers. Referenes Ehrgott, M Multiriteria optimization, volume 2. Springer Berlin. Finkel, R. A., and Bentley, J. L Quad trees a data struture for retrieval on omposite keys. Ata informatia 4(1):1 9. Gavanelli, M An algorithm for multi-riteria optimization in CSPs. ECAI Haeniht, W Quad trees, a datastruture for disrete vetor optimization prolems. In Essays and Surveys on Multiple Criteria Deision Making. Springer Haimes, Y. Y.; Lasdon, L. S.; and Wismer, D. A On a iriterion formulation of the prolems of integrated system identifiation and system optimization. IEEE Transations on Systems, Man, and Cyernetis 1(3): Le Pape, C.; Couronné, P.; Vergamini, D.; and Gosselin, V Time-versus-apaity ompromises in projet sheduling. Lust, T., and Teghem, J Two-phase Pareto loal searh for the iojetive traveling salesman prolem. Journal of Heuristis 16(3): Mavrotas, G Generation of effiient solutions in multiojetive mathematial programming prolems using GAMS. Effetive implementation of the ε-onstraint method. OsaR Team OsaR: Sala in OR. Availale from Paquete, L., and Stützle, T Design and analysis of stohasti loal searh for the multiojetive traveling salesman prolem. Computers & operations researh 36(9): Ralphs, T. K.; Saltzman, M. J.; and Wieek, M. M An improved algorithm for solving iojetive integer programs. Annals of Operations Researh 147(1): Samet, H Foundations of multidimensional and metri data strutures. Morgan Kaufmann. Shaus, P., and Hartert, R Multi-Ojetive Large Neighorhood Searh. In 19th International Conferene on Priniples and Pratie of Constraint Programming. Van Wassenhove, L. N., and Gelders, L. F Solving a iriterion sheduling prolem. European Journal of Operations Researh 4: Xavier Gandileux A olletion of test instanes for multiojetive ominato-

7 rial optimization prolems. Availale from Zitzler, E.; Thiele, L.; Laumanns, M.; Fonsea, C. M.; and da Fonsea, V. G Performane assessment of multiojetive optimizers: An analysis and review. Evolutionary Computation, IEEE Transations on 7(2):

A Support-Based Algorithm for the Bi-Objective Pareto Constraint

A Support-Based Algorithm for the Bi-Objective Pareto Constraint Proeedings of the Twenty-Eighth AAAI Conferene on Artifiial Intelligene A Support-Based Algorithm for the Bi-Ojetive Pareto Constraint Renaud Hartert and Pierre Shaus UCLouvain, ICTEAM, Plae Sainte Bare

More information

Incremental Mining of Partial Periodic Patterns in Time-series Databases

Incremental Mining of Partial Periodic Patterns in Time-series Databases CERIAS Teh Report 2000-03 Inremental Mining of Partial Periodi Patterns in Time-series Dataases Mohamed G. Elfeky Center for Eduation and Researh in Information Assurane and Seurity Purdue University,

More information

A Compressed Breadth-First Search for Satisfiability

A Compressed Breadth-First Search for Satisfiability A Compressed Breadth-First Searh for Satisfiaility DoRon B. Motter and Igor L. Markov Department of EECS, University of Mihigan, 1301 Beal Ave, Ann Aror, MI 48109-2122 dmotter, imarkov @ees.umih.edu Astrat.

More information

Path Sharing and Predicate Evaluation for High-Performance XML Filtering*

Path Sharing and Predicate Evaluation for High-Performance XML Filtering* Path Sharing and Prediate Evaluation for High-Performane XML Filtering Yanlei Diao, Mihael J. Franklin, Hao Zhang, Peter Fisher EECS, University of California, Berkeley {diaoyl, franklin, nhz, fisherp}@s.erkeley.edu

More information

Efficient and scalable trie-based algorithms for computing set containment relations

Efficient and scalable trie-based algorithms for computing set containment relations Effiient and salale trie-ased algorithms for omputing set ontainment relations Yongming Luo #1, George H. L. Flether #2, Jan Hidders 3, Paul De Bra #4 # Eindhoven University of Tehnology, The Netherlands

More information

Algorithms for External Memory Lecture 6 Graph Algorithms - Weighted List Ranking

Algorithms for External Memory Lecture 6 Graph Algorithms - Weighted List Ranking Algorithms for External Memory Leture 6 Graph Algorithms - Weighted List Ranking Leturer: Nodari Sithinava Sribe: Andi Hellmund, Simon Ohsenreither 1 Introdution & Motivation After talking about I/O-effiient

More information

Extracting Partition Statistics from Semistructured Data

Extracting Partition Statistics from Semistructured Data Extrating Partition Statistis from Semistrutured Data John N. Wilson Rihard Gourlay Robert Japp Mathias Neumüller Department of Computer and Information Sienes University of Strathlyde, Glasgow, UK {jnw,rsg,rpj,mathias}@is.strath.a.uk

More information

Learning Convention Propagation in BeerAdvocate Reviews from a etwork Perspective. Abstract

Learning Convention Propagation in BeerAdvocate Reviews from a etwork Perspective. Abstract CS 9 Projet Final Report: Learning Convention Propagation in BeerAdvoate Reviews from a etwork Perspetive Abstrat We look at the way onventions propagate between reviews on the BeerAdvoate dataset, and

More information

Automatic Physical Design Tuning: Workload as a Sequence Sanjay Agrawal Microsoft Research One Microsoft Way Redmond, WA, USA +1-(425)

Automatic Physical Design Tuning: Workload as a Sequence Sanjay Agrawal Microsoft Research One Microsoft Way Redmond, WA, USA +1-(425) Automati Physial Design Tuning: Workload as a Sequene Sanjay Agrawal Mirosoft Researh One Mirosoft Way Redmond, WA, USA +1-(425) 75-357 sagrawal@mirosoft.om Eri Chu * Computer Sienes Department University

More information

A Partial Sorting Algorithm in Multi-Hop Wireless Sensor Networks

A Partial Sorting Algorithm in Multi-Hop Wireless Sensor Networks A Partial Sorting Algorithm in Multi-Hop Wireless Sensor Networks Abouberine Ould Cheikhna Department of Computer Siene University of Piardie Jules Verne 80039 Amiens Frane Ould.heikhna.abouberine @u-piardie.fr

More information

Chapter 2: Introduction to Maple V

Chapter 2: Introduction to Maple V Chapter 2: Introdution to Maple V 2-1 Working with Maple Worksheets Try It! (p. 15) Start a Maple session with an empty worksheet. The name of the worksheet should be Untitled (1). Use one of the standard

More information

HEXA: Compact Data Structures for Faster Packet Processing

HEXA: Compact Data Structures for Faster Packet Processing Washington University in St. Louis Washington University Open Sholarship All Computer Siene and Engineering Researh Computer Siene and Engineering Report Number: 27-26 27 HEXA: Compat Data Strutures for

More information

This fact makes it difficult to evaluate the cost function to be minimized

This fact makes it difficult to evaluate the cost function to be minimized RSOURC LLOCTION N SSINMNT In the resoure alloation step the amount of resoures required to exeute the different types of proesses is determined. We will refer to the time interval during whih a proess

More information

A DYNAMIC ACCESS CONTROL WITH BINARY KEY-PAIR

A DYNAMIC ACCESS CONTROL WITH BINARY KEY-PAIR Malaysian Journal of Computer Siene, Vol 10 No 1, June 1997, pp 36-41 A DYNAMIC ACCESS CONTROL WITH BINARY KEY-PAIR Md Rafiqul Islam, Harihodin Selamat and Mohd Noor Md Sap Faulty of Computer Siene and

More information

Cluster-Based Cumulative Ensembles

Cluster-Based Cumulative Ensembles Cluster-Based Cumulative Ensembles Hanan G. Ayad and Mohamed S. Kamel Pattern Analysis and Mahine Intelligene Lab, Eletrial and Computer Engineering, University of Waterloo, Waterloo, Ontario N2L 3G1,

More information

Dynamic Programming. Lecture #8 of Algorithms, Data structures and Complexity. Joost-Pieter Katoen Formal Methods and Tools Group

Dynamic Programming. Lecture #8 of Algorithms, Data structures and Complexity. Joost-Pieter Katoen Formal Methods and Tools Group Dynami Programming Leture #8 of Algorithms, Data strutures and Complexity Joost-Pieter Katoen Formal Methods and Tools Group E-mail: katoen@s.utwente.nl Otober 29, 2002 JPK #8: Dynami Programming ADC (214020)

More information

Figure 1. LBP in the field of texture analysis operators.

Figure 1. LBP in the field of texture analysis operators. L MEHODOLOGY he loal inary pattern (L) texture analysis operator is defined as a gray-sale invariant texture measure, derived from a general definition of texture in a loal neighorhood. he urrent form

More information

On - Line Path Delay Fault Testing of Omega MINs M. Bellos 1, E. Kalligeros 1, D. Nikolos 1,2 & H. T. Vergos 1,2

On - Line Path Delay Fault Testing of Omega MINs M. Bellos 1, E. Kalligeros 1, D. Nikolos 1,2 & H. T. Vergos 1,2 On - Line Path Delay Fault Testing of Omega MINs M. Bellos, E. Kalligeros, D. Nikolos,2 & H. T. Vergos,2 Dept. of Computer Engineering and Informatis 2 Computer Tehnology Institute University of Patras,

More information

Exploring the Commonality in Feature Modeling Notations

Exploring the Commonality in Feature Modeling Notations Exploring the Commonality in Feature Modeling Notations Miloslav ŠÍPKA Slovak University of Tehnology Faulty of Informatis and Information Tehnologies Ilkovičova 3, 842 16 Bratislava, Slovakia miloslav.sipka@gmail.om

More information

Layout Compliance for Triple Patterning Lithography: An Iterative Approach

Layout Compliance for Triple Patterning Lithography: An Iterative Approach Layout Compliane for Triple Patterning Lithography: An Iterative Approah Bei Yu, Gilda Garreton, David Z. Pan ECE Dept. University of Texas at Austin, Austin, TX, USA Orale Las, Orale Corporation, Redwood

More information

The Minimum Redundancy Maximum Relevance Approach to Building Sparse Support Vector Machines

The Minimum Redundancy Maximum Relevance Approach to Building Sparse Support Vector Machines The Minimum Redundany Maximum Relevane Approah to Building Sparse Support Vetor Mahines Xiaoxing Yang, Ke Tang, and Xin Yao, Nature Inspired Computation and Appliations Laboratory (NICAL), Shool of Computer

More information

A {k, n}-secret Sharing Scheme for Color Images

A {k, n}-secret Sharing Scheme for Color Images A {k, n}-seret Sharing Sheme for Color Images Rastislav Luka, Konstantinos N. Plataniotis, and Anastasios N. Venetsanopoulos The Edward S. Rogers Sr. Dept. of Eletrial and Computer Engineering, University

More information

Circular Pruning for Lazy Diagnosis of Active Systems

Circular Pruning for Lazy Diagnosis of Active Systems Cirular Pruning for Lazy Diagnosis of Ative Systems Andrea Duoli, Gianfrano Lamperti, Emanuele Piantoni, Marina Zanella Dipartimento di Elettronia per l Automazione, Università di Bresia, Italy Astrat:

More information

Mining Edge-Weighted Call Graphs to Localise Software Bugs

Mining Edge-Weighted Call Graphs to Localise Software Bugs Mining Edge-Weighted Call Graphs to Loalise Software Bugs Frank Eihinger, Klemens Böhm, and Matthias Huer Institute for Program Strutures and Data Organisation (IPD), Universität Karlsruhe (TH), Germany

More information

Test Case Generation from UML State Machines

Test Case Generation from UML State Machines Test Case Generation from UML State Mahines Dirk Seifert Loria Université Nany 2 Campus Sientifique, BP 239 F-54506 Vandoeuvre lès Nany edex Dirk.Seifert@Loria.fr inria-00268864, version 2-23 Apr 2008

More information

Algorithms, Mechanisms and Procedures for the Computer-aided Project Generation System

Algorithms, Mechanisms and Procedures for the Computer-aided Project Generation System Algorithms, Mehanisms and Proedures for the Computer-aided Projet Generation System Anton O. Butko 1*, Aleksandr P. Briukhovetskii 2, Dmitry E. Grigoriev 2# and Konstantin S. Kalashnikov 3 1 Department

More information

A Novel Validity Index for Determination of the Optimal Number of Clusters

A Novel Validity Index for Determination of the Optimal Number of Clusters IEICE TRANS. INF. & SYST., VOL.E84 D, NO.2 FEBRUARY 2001 281 LETTER A Novel Validity Index for Determination of the Optimal Number of Clusters Do-Jong KIM, Yong-Woon PARK, and Dong-Jo PARK, Nonmembers

More information

Outline. CS38 Introduction to Algorithms. Administrative Stuff. Administrative Stuff. Motivation/Overview. Administrative Stuff

Outline. CS38 Introduction to Algorithms. Administrative Stuff. Administrative Stuff. Motivation/Overview. Administrative Stuff Outline CS38 Introdution to Algorithms Leture 1 April 1, 2014 administrative stuff motivation and overview of the ourse stale mathings example graphs, representing graphs graph traversals (BFS, DFS) onnetivity,

More information

Machine Vision. Laboratory Exercise Name: Student ID: S

Machine Vision. Laboratory Exercise Name: Student ID: S Mahine Vision 521466S Laoratory Eerise 2011 Name: Student D: General nformation To pass these laoratory works, you should answer all questions (Q.y) with an understandale handwriting either in English

More information

Structural Topology Optimization Based on the Smoothed Finite Element Method

Structural Topology Optimization Based on the Smoothed Finite Element Method 378 Strutural Topology Optimization Based on the Smoothed Finite Element Method Astrat In this paper, the smoothed finite element method, inorporated with the level set method, is employed to arry out

More information

13.1 Numerical Evaluation of Integrals Over One Dimension

13.1 Numerical Evaluation of Integrals Over One Dimension 13.1 Numerial Evaluation of Integrals Over One Dimension A. Purpose This olletion of subprograms estimates the value of the integral b a f(x) dx where the integrand f(x) and the limits a and b are supplied

More information

On Optimal Total Cost and Optimal Order Quantity for Fuzzy Inventory Model without Shortage

On Optimal Total Cost and Optimal Order Quantity for Fuzzy Inventory Model without Shortage International Journal of Fuzzy Mathemat and Systems. ISSN 48-9940 Volume 4, Numer (014, pp. 193-01 Researh India Puliations http://www.ripuliation.om On Optimal Total Cost and Optimal Order Quantity for

More information

Methods for Multi-Dimensional Robustness Optimization in Complex Embedded Systems

Methods for Multi-Dimensional Robustness Optimization in Complex Embedded Systems Methods for Multi-Dimensional Robustness Optimization in Complex Embedded Systems Arne Hamann, Razvan Rau, Rolf Ernst Institute of Computer and Communiation Network Engineering Tehnial University of Braunshweig,

More information

Definitions Homework. Quine McCluskey Optimal solutions are possible for some large functions Espresso heuristic. Definitions Homework

Definitions Homework. Quine McCluskey Optimal solutions are possible for some large functions Espresso heuristic. Definitions Homework EECS 33 There be Dragons here http://ziyang.ees.northwestern.edu/ees33/ Teaher: Offie: Email: Phone: L477 Teh dikrp@northwestern.edu 847 467 2298 Today s material might at first appear diffiult Perhaps

More information

International Journal of Advancements in Research & Technology, Volume 3, Issue 3, March-2014 ISSN

International Journal of Advancements in Research & Technology, Volume 3, Issue 3, March-2014 ISSN International Journal of Advanements in Researh & Tehnology, Volume 3, Issue 3, Marh-204 ISSN 2278-773 47 Phrase Based Doument Retrieving y Comining Suffix Tree index data struture and Boyer- Moore faster

More information

Automated Test Generation from Vulnerability Signatures

Automated Test Generation from Vulnerability Signatures Automated Test Generation from Vulneraility Signatures Adulaki Aydin, Muath Alkhalaf, and Tevfik Bultan Computer Siene Department University of California, Santa Barara Email: {aki,muath,ultan}@s.us.edu

More information

8 : Learning Fully Observed Undirected Graphical Models

8 : Learning Fully Observed Undirected Graphical Models 10-708: Probabilisti Graphial Models 10-708, Spring 2018 8 : Learning Fully Observed Undireted Graphial Models Leturer: Kayhan Batmanghelih Sribes: Chenghui Zhou, Cheng Ran (Harvey) Zhang When learning

More information

Gray Codes for Reflectable Languages

Gray Codes for Reflectable Languages Gray Codes for Refletable Languages Yue Li Joe Sawada Marh 8, 2008 Abstrat We lassify a type of language alled a refletable language. We then develop a generi algorithm that an be used to list all strings

More information

1 Disjoint-set data structure.

1 Disjoint-set data structure. CS 124 Setion #4 Union-Fin, Greey Algorithms 2/20/17 1 Disjoint-set ata struture. 1.1 Operations Disjoint-set ata struture enale us to effiiently perform operations suh as plaing elements into sets, querying

More information

Multi-Piece Mold Design Based on Linear Mixed-Integer Program Toward Guaranteed Optimality

Multi-Piece Mold Design Based on Linear Mixed-Integer Program Toward Guaranteed Optimality INTERNATIONAL CONFERENCE ON MANUFACTURING AUTOMATION (ICMA200) Multi-Piee Mold Design Based on Linear Mixed-Integer Program Toward Guaranteed Optimality Stephen Stoyan, Yong Chen* Epstein Department of

More information

Improved flooding of broadcast messages using extended multipoint relaying

Improved flooding of broadcast messages using extended multipoint relaying Improved flooding of broadast messages using extended multipoint relaying Pere Montolio Aranda a, Joaquin Garia-Alfaro a,b, David Megías a a Universitat Oberta de Catalunya, Estudis d Informàtia, Mulimèdia

More information

Evolutionary Feature Synthesis for Image Databases

Evolutionary Feature Synthesis for Image Databases Evolutionary Feature Synthesis for Image Databases Anlei Dong, Bir Bhanu, Yingqiang Lin Center for Researh in Intelligent Systems University of California, Riverside, California 92521, USA {adong, bhanu,

More information

A Novel Bit Level Time Series Representation with Implication of Similarity Search and Clustering

A Novel Bit Level Time Series Representation with Implication of Similarity Search and Clustering A Novel Bit Level Time Series Representation with Impliation of Similarity Searh and lustering hotirat Ratanamahatana, Eamonn Keogh, Anthony J. Bagnall 2, and Stefano Lonardi Dept. of omputer Siene & Engineering,

More information

Unsupervised Stereoscopic Video Object Segmentation Based on Active Contours and Retrainable Neural Networks

Unsupervised Stereoscopic Video Object Segmentation Based on Active Contours and Retrainable Neural Networks Unsupervised Stereosopi Video Objet Segmentation Based on Ative Contours and Retrainable Neural Networks KLIMIS NTALIANIS, ANASTASIOS DOULAMIS, and NIKOLAOS DOULAMIS National Tehnial University of Athens

More information

Sequential Incremental-Value Auctions

Sequential Incremental-Value Auctions Sequential Inremental-Value Autions Xiaoming Zheng and Sven Koenig Department of Computer Siene University of Southern California Los Angeles, CA 90089-0781 {xiaominz,skoenig}@us.edu Abstrat We study the

More information

NONLINEAR BACK PROJECTION FOR TOMOGRAPHIC IMAGE RECONSTRUCTION. Ken Sauer and Charles A. Bouman

NONLINEAR BACK PROJECTION FOR TOMOGRAPHIC IMAGE RECONSTRUCTION. Ken Sauer and Charles A. Bouman NONLINEAR BACK PROJECTION FOR TOMOGRAPHIC IMAGE RECONSTRUCTION Ken Sauer and Charles A. Bouman Department of Eletrial Engineering, University of Notre Dame Notre Dame, IN 46556, (219) 631-6999 Shool of

More information

Approximate logic synthesis for error tolerant applications

Approximate logic synthesis for error tolerant applications Approximate logi synthesis for error tolerant appliations Doohul Shin and Sandeep K. Gupta Eletrial Engineering Department, University of Southern California, Los Angeles, CA 989 {doohuls, sandeep}@us.edu

More information

High-level synthesis under I/O Timing and Memory constraints

High-level synthesis under I/O Timing and Memory constraints Highlevel synthesis under I/O Timing and Memory onstraints Philippe Coussy, Gwenolé Corre, Pierre Bomel, Eri Senn, Eri Martin To ite this version: Philippe Coussy, Gwenolé Corre, Pierre Bomel, Eri Senn,

More information

Menu. X + /X=1 and XY+X /Y = X(Y + /Y) = X

Menu. X + /X=1 and XY+X /Y = X(Y + /Y) = X Menu K-Maps and Boolean Algera >Don t ares >5 Variale Look into my... 1 Karnaugh Maps - Boolean Algera We have disovered that simplifiation/minimization is an art. If you see it, GREAT! Else, work at it,

More information

Drawing lines. Naïve line drawing algorithm. drawpixel(x, round(y)); double dy = y1 - y0; double dx = x1 - x0; double m = dy / dx; double y = y0;

Drawing lines. Naïve line drawing algorithm. drawpixel(x, round(y)); double dy = y1 - y0; double dx = x1 - x0; double m = dy / dx; double y = y0; Naïve line drawing algorithm // Connet to grid points(x0,y0) and // (x1,y1) by a line. void drawline(int x0, int y0, int x1, int y1) { int x; double dy = y1 - y0; double dx = x1 - x0; double m = dy / dx;

More information

Outline: Software Design

Outline: Software Design Outline: Software Design. Goals History of software design ideas Design priniples Design methods Life belt or leg iron? (Budgen) Copyright Nany Leveson, Sept. 1999 A Little History... At first, struggling

More information

Detecting Outliers in High-Dimensional Datasets with Mixed Attributes

Detecting Outliers in High-Dimensional Datasets with Mixed Attributes Deteting Outliers in High-Dimensional Datasets with Mixed Attributes A. Koufakou, M. Georgiopoulos, and G.C. Anagnostopoulos 2 Shool of EECS, University of Central Florida, Orlando, FL, USA 2 Dept. of

More information

INTEGRATING PHOTOGRAMMETRY AND INERTIAL SENSORS FOR ROBOTICS NAVIGATION AND MAPPING

INTEGRATING PHOTOGRAMMETRY AND INERTIAL SENSORS FOR ROBOTICS NAVIGATION AND MAPPING INTEGRATING PHOTOGRAMMETRY AND INERTIAL SENSORS FOR ROBOTICS NAVIGATION AND MAPPING Fadi Bayoud, Jan Skaloud, Bertrand Merminod Eole Polytehnique Fédérale de Lausanne (EPFL) Geodeti Engineering Laoratory

More information

Implementing Load-Balanced Switches With Fat-Tree Networks

Implementing Load-Balanced Switches With Fat-Tree Networks Implementing Load-Balaned Swithes With Fat-Tree Networks Hung-Shih Chueh, Ching-Min Lien, Cheng-Shang Chang, Jay Cheng, and Duan-Shin Lee Department of Eletrial Engineering & Institute of Communiations

More information

Contents Contents...I List of Tables...VIII List of Figures...IX 1. Introduction Information Retrieval... 8

Contents Contents...I List of Tables...VIII List of Figures...IX 1. Introduction Information Retrieval... 8 Contents Contents...I List of Tables...VIII List of Figures...IX 1. Introdution... 1 1.1. Internet Information...2 1.2. Internet Information Retrieval...3 1.2.1. Doument Indexing...4 1.2.2. Doument Retrieval...4

More information

An Efficient and Scalable Approach to CNN Queries in a Road Network

An Efficient and Scalable Approach to CNN Queries in a Road Network An Effiient and Salable Approah to CNN Queries in a Road Network Hyung-Ju Cho Chin-Wan Chung Dept. of Eletrial Engineering & Computer Siene Korea Advaned Institute of Siene and Tehnology 373- Kusong-dong,

More information

COMBINATION OF INTERSECTION- AND SWEPT-BASED METHODS FOR SINGLE-MATERIAL REMAP

COMBINATION OF INTERSECTION- AND SWEPT-BASED METHODS FOR SINGLE-MATERIAL REMAP Combination of intersetion- and swept-based methods for single-material remap 11th World Congress on Computational Mehanis WCCM XI) 5th European Conferene on Computational Mehanis ECCM V) 6th European

More information

Semi-Supervised Affinity Propagation with Instance-Level Constraints

Semi-Supervised Affinity Propagation with Instance-Level Constraints Semi-Supervised Affinity Propagation with Instane-Level Constraints Inmar E. Givoni, Brendan J. Frey Probabilisti and Statistial Inferene Group University of Toronto 10 King s College Road, Toronto, Ontario,

More information

One Against One or One Against All : Which One is Better for Handwriting Recognition with SVMs?

One Against One or One Against All : Which One is Better for Handwriting Recognition with SVMs? One Against One or One Against All : Whih One is Better for Handwriting Reognition with SVMs? Jonathan Milgram, Mohamed Cheriet, Robert Sabourin To ite this version: Jonathan Milgram, Mohamed Cheriet,

More information

Facility Location: Distributed Approximation

Facility Location: Distributed Approximation Faility Loation: Distributed Approximation Thomas Mosibroda Roger Wattenhofer Distributed Computing Group PODC 2005 Where to plae ahes in the Internet? A distributed appliation that has to dynamially plae

More information

Multi-Level Modeling of Concurrent and Distributed Systems

Multi-Level Modeling of Concurrent and Distributed Systems Multi-Level Modeling of Conurrent and Distriuted Systems Peter Taeling Hasso-Plattner-Institute for Software Systems Engineering P.O. Box 90 04 60, 14440 Potsdam, Germany taeling@hpi.uni-potsdam.de strat

More information

Multiple-Criteria Decision Analysis: A Novel Rank Aggregation Method

Multiple-Criteria Decision Analysis: A Novel Rank Aggregation Method 3537 Multiple-Criteria Deision Analysis: A Novel Rank Aggregation Method Derya Yiltas-Kaplan Department of Computer Engineering, Istanbul University, 34320, Avilar, Istanbul, Turkey Email: dyiltas@ istanbul.edu.tr

More information

We don t need no generation - a practical approach to sliding window RLNC

We don t need no generation - a practical approach to sliding window RLNC We don t need no generation - a pratial approah to sliding window RLNC Simon Wunderlih, Frank Gabriel, Sreekrishna Pandi, Frank H.P. Fitzek Deutshe Telekom Chair of Communiation Networks, TU Dresden, Dresden,

More information

An Interactive-Voting Based Map Matching Algorithm

An Interactive-Voting Based Map Matching Algorithm Eleventh International Conferene on Mobile Data Management An Interative-Voting Based Map Mathing Algorithm Jing Yuan* University of Siene and Tehnology of China Hefei, China yuanjing@mail.ust.edu.n Yu

More information

The Implementation of RRTs for a Remote-Controlled Mobile Robot

The Implementation of RRTs for a Remote-Controlled Mobile Robot ICCAS5 June -5, KINEX, Gyeonggi-Do, Korea he Implementation of RRs for a Remote-Controlled Mobile Robot Chi-Won Roh*, Woo-Sub Lee **, Sung-Chul Kang *** and Kwang-Won Lee **** * Intelligent Robotis Researh

More information

Sparse Certificates for 2-Connectivity in Directed Graphs

Sparse Certificates for 2-Connectivity in Directed Graphs Sparse Certifiates for 2-Connetivity in Direted Graphs Loukas Georgiadis Giuseppe F. Italiano Aikaterini Karanasiou Charis Papadopoulos Nikos Parotsidis Abstrat Motivated by the emergene of large-sale

More information

Analysis of input and output configurations for use in four-valued CCD programmable logic arrays

Analysis of input and output configurations for use in four-valued CCD programmable logic arrays nalysis of input and output onfigurations for use in four-valued D programmable logi arrays J.T. utler H.G. Kerkhoff ndexing terms: Logi, iruit theory and design, harge-oupled devies bstrat: s in binary,

More information

Abstract. Key Words: Image Filters, Fuzzy Filters, Order Statistics Filters, Rank Ordered Mean Filters, Channel Noise. 1.

Abstract. Key Words: Image Filters, Fuzzy Filters, Order Statistics Filters, Rank Ordered Mean Filters, Channel Noise. 1. Fuzzy Weighted Rank Ordered Mean (FWROM) Filters for Mixed Noise Suppression from Images S. Meher, G. Panda, B. Majhi 3, M.R. Meher 4,,4 Department of Eletronis and I.E., National Institute of Tehnology,

More information

Naïve Bayesian Rough Sets Under Fuzziness

Naïve Bayesian Rough Sets Under Fuzziness IJMSA: Vol. 6, No. 1-2, January-June 2012, pp. 19 25 Serials Publiations ISSN: 0973-6786 Naïve ayesian Rough Sets Under Fuzziness G. GANSAN 1,. KRISHNAVNI 2 T. HYMAVATHI 3 1,2,3 Department of Mathematis,

More information

System-Level Parallelism and Throughput Optimization in Designing Reconfigurable Computing Applications

System-Level Parallelism and Throughput Optimization in Designing Reconfigurable Computing Applications System-Level Parallelism and hroughput Optimization in Designing Reonfigurable Computing Appliations Esam El-Araby 1, Mohamed aher 1, Kris Gaj 2, arek El-Ghazawi 1, David Caliga 3, and Nikitas Alexandridis

More information

Capturing Large Intra-class Variations of Biometric Data by Template Co-updating

Capturing Large Intra-class Variations of Biometric Data by Template Co-updating Capturing Large Intra-lass Variations of Biometri Data by Template Co-updating Ajita Rattani University of Cagliari Piazza d'armi, Cagliari, Italy ajita.rattani@diee.unia.it Gian Lua Marialis University

More information

Abstract. We describe a parametric hybrid Bezier patch that, in addition. schemes are local in that changes to part of the data only aect portions of

Abstract. We describe a parametric hybrid Bezier patch that, in addition. schemes are local in that changes to part of the data only aect portions of A Parametri Hyrid Triangular Bezier Path Stephen Mann and Matthew Davidhuk Astrat. We desrie a parametri hyrid Bezier path that, in addition to lending interior ontrol points, lends oundary ontrol points.

More information

Partial Character Decoding for Improved Regular Expression Matching in FPGAs

Partial Character Decoding for Improved Regular Expression Matching in FPGAs Partial Charater Deoding for Improved Regular Expression Mathing in FPGAs Peter Sutton Shool of Information Tehnology and Eletrial Engineering The University of Queensland Brisbane, Queensland, 4072, Australia

More information

1. The collection of the vowels in the word probability. 2. The collection of real numbers that satisfy the equation x 9 = 0.

1. The collection of the vowels in the word probability. 2. The collection of real numbers that satisfy the equation x 9 = 0. C HPTER 1 SETS I. DEFINITION OF SET We begin our study of probability with the disussion of the basi onept of set. We assume that there is a ommon understanding of what is meant by the notion of a olletion

More information

Scheduling Multiple Independent Hard-Real-Time Jobs on a Heterogeneous Multiprocessor

Scheduling Multiple Independent Hard-Real-Time Jobs on a Heterogeneous Multiprocessor Sheduling Multiple Independent Hard-Real-Time Jobs on a Heterogeneous Multiproessor Orlando Moreira NXP Semiondutors Researh Eindhoven, Netherlands orlando.moreira@nxp.om Frederio Valente Universidade

More information

A Hybrid Neuro-Genetic Approach to Short-Term Traffic Volume Prediction

A Hybrid Neuro-Genetic Approach to Short-Term Traffic Volume Prediction A Hybrid Neuro-Geneti Approah to Short-Term Traffi Volume Predition 1. Introdution Shahriar Afandizadeh 1,*, Jalil Kianfar 2 Reeived: January 2003, Revised: July 2008, Aepted: January 2009 Abstrat: This

More information

Gradient based progressive probabilistic Hough transform

Gradient based progressive probabilistic Hough transform Gradient based progressive probabilisti Hough transform C.Galambos, J.Kittler and J.Matas Abstrat: The authors look at the benefits of exploiting gradient information to enhane the progressive probabilisti

More information

AN ANT BASED SIMULATION OPTIMIZATION FOR VEHICLE ROUTING PROBLEM WITH STOCHASTIC DEMANDS. Mukul Tripathi Glenn Kuriger Hung-da Wan

AN ANT BASED SIMULATION OPTIMIZATION FOR VEHICLE ROUTING PROBLEM WITH STOCHASTIC DEMANDS. Mukul Tripathi Glenn Kuriger Hung-da Wan Proeedings of the 29 Winter Simulation Conferene M. D. Rossetti, R. R. Hill, B. Johansson, A. Dunin and R. G. Ingalls, eds. AN ANT BASED SIMULATION OPTIMIZATION FOR VEHICLE ROUTING PROBLEM WITH STOCHASTIC

More information

Z Combinatorial Filters: Sensor Beams, Obstacles, and Possible Paths

Z Combinatorial Filters: Sensor Beams, Obstacles, and Possible Paths Z Combinatorial Filters: Sensor Beams, Obstales, and Possible Paths BENJAMIN TOVAR, Northwestern University FRED COHEN, University of Rohester LEONARDO BOBADILLA, University of Illinois JUSTIN CZARNOWSKI,

More information

the data. Structured Principal Component Analysis (SPCA)

the data. Structured Principal Component Analysis (SPCA) Strutured Prinipal Component Analysis Kristin M. Branson and Sameer Agarwal Department of Computer Siene and Engineering University of California, San Diego La Jolla, CA 9193-114 Abstrat Many tasks involving

More information

A Load-Balanced Clustering Protocol for Hierarchical Wireless Sensor Networks

A Load-Balanced Clustering Protocol for Hierarchical Wireless Sensor Networks International Journal of Advanes in Computer Networks and Its Seurity IJCNS A Load-Balaned Clustering Protool for Hierarhial Wireless Sensor Networks Mehdi Tarhani, Yousef S. Kavian, Saman Siavoshi, Ali

More information

Relevance for Computer Vision

Relevance for Computer Vision The Geometry of ROC Spae: Understanding Mahine Learning Metris through ROC Isometris, by Peter A. Flah International Conferene on Mahine Learning (ICML-23) http://www.s.bris.a.uk/publiations/papers/74.pdf

More information

Beyond Procedural Facade Parsing: Bidirectional Alignment via Linear Programming

Beyond Procedural Facade Parsing: Bidirectional Alignment via Linear Programming Beyond Proedural Faade Parsing: Bidiretional Alignment via Linear Programming Mateusz Koziński, Guillaume Oozinski and Renaud Marlet Université Paris-Est, LIGM (UMR CNRS 8049), ENPC F-77455 Marne-la-Vallée

More information

Parallelizing Frequent Web Access Pattern Mining with Partial Enumeration for High Speedup

Parallelizing Frequent Web Access Pattern Mining with Partial Enumeration for High Speedup Parallelizing Frequent Web Aess Pattern Mining with Partial Enumeration for High Peiyi Tang Markus P. Turkia Department of Computer Siene Department of Computer Siene University of Arkansas at Little Rok

More information

Efficient Implementation of Beam-Search Incremental Parsers

Efficient Implementation of Beam-Search Incremental Parsers Effiient Implementation of Beam-Searh Inremental Parsers Yoav Goldberg Dept. of Computer Siene Bar-Ilan University Ramat Gan, Tel Aviv, 5290002 Israel yoav.goldberg@gmail.om Kai Zhao Liang Huang Graduate

More information

Routing Protocols for Wireless Ad Hoc Networks Hybrid routing protocols Theofanis Kilinkaridis

Routing Protocols for Wireless Ad Hoc Networks Hybrid routing protocols Theofanis Kilinkaridis Routing Protools for Wireless Ad Ho Networks Hyrid routing protools Theofanis Kilinkaridis tkilinka@.hut.fi Astrat This paper presents a partiular group of routing protools that aim to omine the advantages

More information

3-D IMAGE MODELS AND COMPRESSION - SYNTHETIC HYBRID OR NATURAL FIT?

3-D IMAGE MODELS AND COMPRESSION - SYNTHETIC HYBRID OR NATURAL FIT? 3-D IMAGE MODELS AND COMPRESSION - SYNTHETIC HYBRID OR NATURAL FIT? Bernd Girod, Peter Eisert, Marus Magnor, Ekehard Steinbah, Thomas Wiegand Te {girod eommuniations Laboratory, University of Erlangen-Nuremberg

More information

Locality-Sensitive Hashing Scheme Based on p-stable Distributions

Locality-Sensitive Hashing Scheme Based on p-stable Distributions Loality-Sensitive Hashing Sheme Based on p-stable Distributions Mayur Datar Department of Computer Siene, Stanford University datar@s.stanford.edu Piotr Indyk Laboratory for Computer Siene, MIT indyk@theory.ls.mit.edu

More information

Exploiting Longer Cycles for Link Prediction in Signed Networks

Exploiting Longer Cycles for Link Prediction in Signed Networks Exploiting Longer Cyles for Link Predition in Signed Networks Kai-Yang Chiang kyhiang@s.utexas.edu Nagarajan Natarajan naga86@s.utexas.edu Inderjit S. Dhillon inderjit@s.utexas.edu Amuj Tewari amuj@s.utexas.edu

More information

Self-Adaptive Parent to Mean-Centric Recombination for Real-Parameter Optimization

Self-Adaptive Parent to Mean-Centric Recombination for Real-Parameter Optimization Self-Adaptive Parent to Mean-Centri Reombination for Real-Parameter Optimization Kalyanmoy Deb and Himanshu Jain Department of Mehanial Engineering Indian Institute of Tehnology Kanpur Kanpur, PIN 86 {deb,hjain}@iitk.a.in

More information

A Dual-Hamiltonian-Path-Based Multicasting Strategy for Wormhole-Routed Star Graph Interconnection Networks

A Dual-Hamiltonian-Path-Based Multicasting Strategy for Wormhole-Routed Star Graph Interconnection Networks A Dual-Hamiltonian-Path-Based Multiasting Strategy for Wormhole-Routed Star Graph Interonnetion Networks Nen-Chung Wang Department of Information and Communiation Engineering Chaoyang University of Tehnology,

More information

RAC 2 E: Novel Rendezvous Protocol for Asynchronous Cognitive Radios in Cooperative Environments

RAC 2 E: Novel Rendezvous Protocol for Asynchronous Cognitive Radios in Cooperative Environments 21st Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communiations 1 RAC 2 E: Novel Rendezvous Protool for Asynhronous Cognitive Radios in Cooperative Environments Valentina Pavlovska,

More information

Automated System for the Study of Environmental Loads Applied to Production Risers Dustin M. Brandt 1, Celso K. Morooka 2, Ivan R.

Automated System for the Study of Environmental Loads Applied to Production Risers Dustin M. Brandt 1, Celso K. Morooka 2, Ivan R. EngOpt 2008 - International Conferene on Engineering Optimization Rio de Janeiro, Brazil, 01-05 June 2008. Automated System for the Study of Environmental Loads Applied to Prodution Risers Dustin M. Brandt

More information

Make your process world

Make your process world Automation platforms Modion Quantum Safety System Make your proess world a safer plae You are faing omplex hallenges... Safety is at the heart of your proess In order to maintain and inrease your ompetitiveness,

More information

\A CENTRAL FACILITIES LOCATION PROBLEM INVOLVING TRAVELING SALESMAN TOURS AND EXPECTED DISTANCES/ MASTER OF SCIENCE

\A CENTRAL FACILITIES LOCATION PROBLEM INVOLVING TRAVELING SALESMAN TOURS AND EXPECTED DISTANCES/ MASTER OF SCIENCE \A CENTRAL FACILITIES LOCATION PROBLEM INVOLVING TRAVELING SALESMAN TOURS AND EXPECTED DISTANCES/ by Thesis submitted to the Graduate Faulty of the Virginia Polytehni Institute and State University in

More information

Time delay estimation of reverberant meeting speech: on the use of multichannel linear prediction

Time delay estimation of reverberant meeting speech: on the use of multichannel linear prediction University of Wollongong Researh Online Faulty of Informatis - apers (Arhive) Faulty of Engineering and Information Sienes 7 Time delay estimation of reverberant meeting speeh: on the use of multihannel

More information

Using Game Theory and Bayesian Networks to Optimize Cooperation in Ad Hoc Wireless Networks

Using Game Theory and Bayesian Networks to Optimize Cooperation in Ad Hoc Wireless Networks Using Game Theory and Bayesian Networks to Optimize Cooperation in Ad Ho Wireless Networks Giorgio Quer, Federio Librino, Lua Canzian, Leonardo Badia, Mihele Zorzi, University of California San Diego La

More information

Stochastic Constraint Programming

Stochastic Constraint Programming Stohasti Constraint Programming Toby Walsh Abstrat. To model ombinatorial deision problems involving unertainty and probability, we introdue stohasti onstraint programming. Stohasti onstraint programs

More information

Calculation of typical running time of a branch-and-bound algorithm for the vertex-cover problem

Calculation of typical running time of a branch-and-bound algorithm for the vertex-cover problem Calulation of typial running time of a branh-and-bound algorithm for the vertex-over problem Joni Pajarinen, Joni.Pajarinen@iki.fi Otober 21, 2007 1 Introdution The vertex-over problem is one of a olletion

More information

Pipelined Multipliers for Reconfigurable Hardware

Pipelined Multipliers for Reconfigurable Hardware Pipelined Multipliers for Reonfigurable Hardware Mithell J. Myjak and José G. Delgado-Frias Shool of Eletrial Engineering and Computer Siene, Washington State University Pullman, WA 99164-2752 USA {mmyjak,

More information