An Action Model Learning Method for Planning in Incomplete STRIPS Domains
|
|
- Matthew Morris
- 6 years ago
- Views:
Transcription
1 An Action Model Learning Method for Planning in Incomplete STRIPS Domains Romulo de O. Leite Volmir E. Wilhelm Departamento de Matemática Universidade Federal do Paraná Curitiba, Brasil Abstract The dependence on complete STRIPS domain models limits the applicability of much of planning technologies. Under incompleteness hypothesis, some knowledge discovery technique is required to aid a planner in the task of building robust plans. This paper presents an Action Model Learning method that consists on overestimating domain s knowledge and posteriorly eliminating eventual inconsistencies based on correct representation rules, in order to discover improved domain models to be delivered for a complete-knowledge-based planner. An example illustrates how the method is applied to a STRIPS planning problem. Keywords automated planning; action model learning; incomplete domains I. INTRODUCTION The problem of planning based on incomplete STRIPS domain models consists on building plans that are solutions for planning problems even when the set of preconditions and effects of the operators are not completely known or stated. This problem arises from the toughness of building complete domains from scratch, which is usually a non-trivial and timeconsuming task. Most work on planning relies on domain completeness. Because of that, current planning techniques show a lack of robustness when facing incompleteness, i.e., have low accuracy in building plans that are effective solutions in a complete context. Works like [1], [2] and [5] suggest that such dependence on completeness represents a bottleneck for the applications of new planning technologies. When completeness cannot be guaranteed, an auxiliary source of information is needed to provide complementary knowledge about the domain. In Case-Based Planning, such source of information consists on a set of solved planning problems that are known to be valid for the complete domain. The main objective of the present work is to propose an action model learning method for gathering knowledge from a set of plan cases, in order to enhance the knowledge about a given STRIPS domain model. This is done by, in a first step, overestimating preconditions and effects of the operators when those are not stated. After that, a sequence of reasonableness rules is applied to eliminate practical and logical inconsistencies. The output is a domain model such that, for all operators, the lists of preconditions and effects contain all literals that are valid for the complete domain model, with possible overkills which are eventually not eliminated based on the information gathered from the set of plan cases. The present paper is organized as follows: section II makes a brief description of related work; section III is concerned about the formal definition for the problem of planning in incomplete domains; sections IV and V present the proposed method s algorithm and an experiment applying it to a planning problem in blocks world domain, respectively; and, at last, section VI contains a brief discussion about results and conclusions from the experiment. II. RELATED WORK There are two main approaches for the planning problem under incompleteness assumption; both apply principles of Case-Based Planning, i.e. make use of a set of correctly solved problems to obtain complementary information about the domain. One is Model-Lite approach, proposed by Kambhampati [6], which is concerned about building plans even when the input domain model is assumed to be incomplete and mining for additional knowledge from the set of plan traces. More on this approach can be found in [8], [10] and [13]. Another approach is Action Model Learning, focused on learning characteristics of the operators in the domain model for the application of some complete-knowledge-based planning technique, making use of available valid plan traces. Zettlemoyer, Pasula and Kaelbling [12] presented an action learning algorithm to deal with noisy and stochastic domains where states are fully observable; ARMS algorithm, proposed by Yang, Wu and Jiang [11] gathers knowledge on the statistical distribution of frequent set of actions in a set of plan traces to build a domain model from scratch; Walsh and Littman [9] addressed web-service composition problem using
2 a raw experience algorithm that constructs pessimistic action models in a procedure that is somewhat similar to the inflation phase in our proposed method; Creswell, McCluskey and West [3] presented the LOCM system to learn action models based only on the set of cases, without any partial domain as input, but requires a relatively large training to be efficient; LAMP algorithm [15] deals with action models with quantifiers and logical implications, but, just like ARMS, tries to build action models from scratch; RIM method use plan traces to learn micro-actions to refine incomplete domain models[14]. In this context, the present work brings up an action model learning method that aims to turn an incomplete model into an overcomplete model, the Domain Inflation and Reduction (DIR) method. Assuming that the lack of knowledge leads to non-accurate plans, the research question proposed is whether using the set of plan traces to eliminate exceeding literals from action models, instead of using it to compose those action models starting from incomplete ones, leads to more efficient learning and, consequently, more accurate solutions for planning problems on incomplete domain models. III. PROBLEM DEFINITION Let L be a function-free first-order language with finitely many predicate symbols and finitely many constant symbols; a STRIPS domain on L is defined as a restricted state-transition system = (S,A, ), such that S is a finite set of STRIPS states, A is a set of ground instances of some STRIPS planning operators, and is the transition function [4]. Note that, in a restricted STRIPS domain as just defined, the current state remains unchanged until the application of some action in A, that is, no events occur in such domain. Let O be the set of operators in ; an operator o O is defined as a 3-tuple (name(o), pre(o), eff(o)), where name(o) is the operator s name, pre(o) is the set of preconditions of o and eff(o) is the set of effects of o. The name of the operator o, name(o), is a syntactic expression of the form n(x 1,, x n ), where n is a unique symbol and x 1,, x n are all the argument variables to o; pre(o) and eff(o) are sets of literals. The set of effects of o has two disjoint subsets: eff + (o) is the set of positive literals in eff(o), also called the add list, and eff - (o) is the set of negative literal, also called the delete list. Only variables that are arguments of o are allowed to figure in the literals in pre(o) and eff(o). In essence, a STRIPS domain is described by its set of operators; so, the completeness of a domain depends on the completeness of its operators. An operator is said to be complete if there are no missing literal in its sets of preconditions and effects; in opposite, an operator is incomplete if there are some missing literals in those sets. Consequently, a domain that contains some incomplete operator is said to be an incomplete domain [13]. A STRIPS planning problem P is a 3-tuple (, s 0, g), such that is a STRIPS domain model, s 0 S is an initial state and g is a goal. A state in S is a set of ground atoms; so, s 0 is described by the set of ground atoms that are assumed to hold at the starting point. On the other hand, g is a set of literals that must hold at the final point. The goal g is satisfied by any state in s such that all literals in g hold. A solution to a STRIPS planning problem is a plan = {a 1,, a k }, where a 1,, a k is an ordered sequence of ground instances of the operators in O, i.e., a i A for i = 1,..., k. The definitions given so far in this section allow to formally define the incomplete domain planning problem: given an incomplete domain *, a planning problem P = ( *, s 0, g) and a set C of successful plan cases with respect to the complete domain, the objective is to find a solution for P that is correct with respect to. An example of incomplete STRIPS domain for blocks world is proposed by Zhuo, Nguyen and Kambhampati [13], as Table I shows. Underlined literals are those omitted from the original complete domain model. In addition, Table I contains the initial state and the goal for a specific planning problem and also shows the set of plan cases. TABLE I. AN EXAMPLE OF INCOMPLETE DOMAIN PLANNING PROBLEM Operator Preconditions Effects pickup (?x not () not () putdown (?x not () unstack (?x?y stack (?x?y Initial state (s 0 ) ontable D on C A Problem ID Initial state (P i ) (s 0 ) P 1 ontable D P 2 on C B Goal (g) on D A on A C on C B on C A on A B not () not () not () not () Goal (g) on B A on C B on D C Plan trace ( i ) {pickup C, stack C B, pickup A, stack A C, pickup D, stack D A} {unstack C B, putdown C, pickup A, stack A B, pickup C, stack C A}
3 IV. THE ALGORITHM In essence, DIR algorithm works in two phases. The first phase aims to inflate the lists of preconditions and effect of all operators in the domain, filling them up with all literals that are candidates to be missing ones. In a first step, the sets of effects of every operator is filled up and, in a second step, every plan trace is used to guide the inflation of the sets of preconditions. The given preconditions and effects are referred as fixed preconditions/effects, whereas the ones added on inflation phase are referred as inflated preconditions/effects. For every solved planning problems in C, starting from the respective initial state and assuming that it meets all preconditions for the application of the first action in the plan, the algorithm parses every predicate that has at most the same number of parameters as the current action; if there exist some literal stating the predicate with respect to some argument in the current action and if such literal is not a ground fixed precondition, the respective unground precondition is added to the set of preconditions of the current operator. After that, the current state is modified through the application of the inflated effects of the current operator. The inflation applied in the first phase produces a set of overfilled operators. However, under the hypothesis of finite predicate and variable symbols, it is possible to state that the complete sets of preconditions and effects are subsets of the sets in those operators on this stage. The inflation phase pseudocode is described by the Algorithm 1 (Table II). Inflation phase is critical in terms of complexity because of the construction of all possible candidates to be missing literals. Let A be an operator in a domain with t types of objects; the set of parameters of A may be composed of objects of s types, 0 s t. Hence, candidate literals are related to predicates which have at most s types of objects. In the worst case scenario, where s = t and all predicates has parameters of t different types, the number c(a) of candidate literals for an operator A is equal to Algorithm 1 TABLE II. INFLATION PHASE ALGORITHM Input: ( *, C) Output: an inflated domain model inf 0: inf = * 1: for every operator o in O do 2: add all the literals that are supposed to be missing to the set of inflated effects of o 3: end for 4: for every plan case P i in C do 5: current state = s 0 (P i ) 6: for every action in i do 7: add all literals that are supposed to be missing to the set of inflated preconditions of o 8: current state = (current state, current action) 9: end for 10: end for 11: return inf The first reduction step imposes constraints to operators individually. According to Yang, Wu and Jiang [11], for an operator to be correctly represented the intersection of its set of preconditions and its set of positive effects must be the empty set. Hence, under the assumption that all the knowledge contained in the given incomplete domain model actually stands for the complete domain, the first reduction step eliminates all inflated preconditions which, when compared to fixed effects, violate such constraint; the same is applied to inflated effects related to fixed preconditions. The first reduction pseudocode is described by Algorithm 2 (Table III). The second reduction step uses plan traces in C to detect inconsistencies in 1. It aims to eliminate inflated preconditions and effects that are not according to the order actions are applied in every trace, if 1 is assumed to be true. If a preceding action a i produces some effect l over some subset of objects in the list of parameters, then the negation of that effect, l, must not be a precondition of a subsequent action a i + 1, and vice-versa; so, when such situation is found, l or l must be eliminated. t i c(a) 4k m i 1 where k is the number of predicates in the domain model (multiplied by 4, because of the inflations of both sets of preconditions and effects with positive and negative literals), m i is the number of objects of type i in the set of parameter of A and n i is the number of objects of type i in the set of parameters of the predicates. Although theoretically possible, such scenario is extremely pessimistic in practice. The second phase is called the reduction phase and it is composed of at least three steps. These steps consist on a systematic application of some rules to eliminate exceeding literals which were added on the first phase. At least means that there are three steps that are applicable independently of the domain; for each case, some domain-specific rules may be applied for a more efficient reduction. n i TABLE III. Algorithm 2 FIRST STEP ON REDUCTION PHASE ALGORITHM Input: inf Output: a partially inflated domain 1 0: 1 = inf 1: for every operator o in O do 2: if a literal l is a fixed precondition and is an inflated effect of o then 3: eliminate l from the set of effects of o 4: else if a literal l is a fixed effect and an inflated precondition of o Then 5: eliminate l from the set of preconditions of o 6: end if 7: end for 8: return 1 The way the inflated domain inf is constructed may yield some contradictions into the sets of preconditions and effects of an operator. For example, l and l may appear in a same set in
4 1. In order to get rid of these contradictions, the second reduction step of DIR algorithm parses the connections between two consecutive actions and eliminates the literals that cause inconsistency. The second reduction step pseudocode is described by Algorithm 3 (Table IV). TABLE IV. Algorithm 3 SECOND STEP ON REDUCTION PHASE ALGORITHM Input: 1 Output: a partially inflated domain 2 0: 2 = 1 1: for every action plan trace j in C do 2: for every pair of consecutive actions a i and a i + 1 in j do 3: if (l is an effect of a i and l is a precondition of a i + 1 ) and ( l is an effect of a i and l is not a precondition of a i + 1 ) then 4: eliminate l from the set of effects of o i a 5: else if (l is an effect of a i and l is a precondition of a i + 1 ) and ( l is not an effect of a i and l is a precondition of a i + 1 ) then 6: eliminate l from the set of preconditions of o i + 1 a 7: end if 8: end for 9: end for 10: return 2 a. oi and o i + 1 are the operators whose a i and a i + 1 are ground instances, respectively. The objective of the third reduction step is to eliminate contradictions that eventually remain in 2. In the inflation phase, absent predicates are candidates to be missing; however, this may not be the case for some predicate. There are three possibilities: a literal l is indeed a missing one or; l is the actual missing literal or; the predicate to what l refers is not a missing predicate. So, if an eventual contradiction remains at this point, the algorithm discards both l and l, assuming that the predicate does not make part of the sets of preconditions or effects of the operator. The third reduction step pseudocode is described by Algorithm 4 (Table V). The three reduction steps presented so far are applicable to any domain model that fits the hypothesis described in section III. An additional reduction step may be applied to provide a better refinement of the domain model; however, some domain-specific knowledge is needed. The next section presents an example of application of the proposed algorithm in a planning problem on the blocks world domain where a fourth reduction step is applied in order to eliminate inflated literals that can be deduced from fixed ones. V. EXPERIMENT Given the problem described in section III, the DIR algorithm begins applying the inflation phase over all operators in the domain. The operator pickup (?x), for example, has three literals in the set of effects: not (), and not (). So, there is no statement about the predicates clear and on. For the predicate clear, there are two candidates to be missing literals: and ; so, both are added to the set of inflated effects of pickup (?x). On the other hand, the predicate on has more arguments than the operator pickup (?x); so, there is no candidate to be missing literals with respect to this predicate in the current operator. The same is applied to the remaining operators: putdown (?x), unstack (?x?y) and stack (?x?y). TABLE V. Algorithm 4 THIRD STEP ON REDUCTION PHASE ALGORITHM Input: 2 Output: a partially inflated domain 3 0: 3 = 2 1: for every operator o in O do 2: if there exist some literal l such that l and l are inflated preconditions then 3: eliminate l and l from the set of preconditions of o 4: else if there exist some literal l such that l and l are inflated effects then 5: eliminate l and l from the set of effects of o 6: end if 7: end for 8: return 3 The initial state s 0 in P 1 meets all the preconditions for the application of the first action in 1 (pickup (C)). Searching for missing preconditions in pickup (?x), it s found that: holds for s 0 and is not a fixed precondition add to the set of inflated preconditions of pickup (?x); holds for s 0 and it is already a fixed precondition; the predicate on does not apply, because it has more arguments than the operator pickup (?x); holding C does not hold in s 0 ; holds for s 0 and it is already a fixed precondition. The application of the inflated action pickup (C) in s 0 leads to another partially observable state s 1 such that, compared to the fully observable state s 1 which would be obtained if the complete domain were given instead *, contains all the literals that should hold in s 1. However, there are eventually literals in s 1 that should not hold in s 1. The transition from s 0 to s 1 is described by Table VI. TABLE VI. EXAMPLE OF TRANSITION WITH AN INFLATED ACTION s 0 pickup (C) s 1 ontable D - fixed preconditions: - inflated preconditions: - fixed effects: not holding C not - inflated effects: not not holding C
5 TABLE VII. putdown (?x EXAMPLE OF DOMAIN AFTER THE INFLATION PHASE Operator Preconditions Effects not () Fixed pickup (?x not () Fixed unstack (?x?y stack (?x?y Fixed Fixed not () not () Handempty not () Handempty not () not () not () not () holding?y not (holding?y) not () not () holding?y not (holding?y) not () not () not () holding?y not (holding?y) After the inflation phase, inf will be as shown on Table VII. Note that there are many antagonistic literals; for example, the set of effects of pickup (?x) has and not () as elements. However, both must not be valid simultaneously for the correct action model. So, the reduction phase algorithm takes inf as input and aims to eliminate such inconsistencies. The first step on reduction phase is to eliminate all inflated literals which are in the intersection of fixed preconditions and inflated effects and also in the intersection of inflated preconditions and fixed effects. For example, the operator putdown (?x) has as an inflated precondition and a fixed effect; hence, it must be eliminated from the set of preconditions. The second step algorithm searches for inconsistencies in the sequences of actions in the plan cases. For example, pickup (C) and stack (C B) are actions a 1 and a 2 in the plan 1, respectively. The literals and are effects of pickup (?x); on the other hand, only is a precondition of stack (?x?y). This leads to the conclusion that must not be an effect of pickup (?x); otherwise, stack (C B) would not be applicable right after pickup (C). Hence, must be eliminated from the set of effects of pickup (?x). TABLE VIII. RESULTING DOMAIN MODEL Operator Preconditions Effects not () pickup (?x not () putdown (?x unstack (?x?y stack (?x?y not () b not () not () not () b not () not () b. Differs from the complete domain model The third step is to eliminate remaining inconsistencies inside the set of inflated preconditions and the set of inflated effects in each operator. For example, the set of inflated effects of the operator unstack (?x?y) remains with both literals and not (). The rule at this point is to assume that eventual remaining inconsistencies indicates that the referred predicate is not likely to be present in the set of effects of unstack (?x?y) in the complete domain and eliminate both literals. So, and not () are both discarded. In addition a fourth reduction step is applied. It consists on the elimination of unnecessary literals, based on domainspecific axioms: for every operator, if all fixed preconditions imply an inflated precondition, this must be eliminated; on the other hand, if all fixed effects imply an inflated effect, this must be eliminated. For example, the operator putdown (?x) has as fixed precondition and and not () as inflated preconditions. It is known from the domain axioms that if holding x is true for some block x, it implies that is also true; the same can be said about not (). So, these two negative preconditions are not necessary, since imply both. After the application of the four reduction steps, the resulting domain model is that described by Table VIII. The resulting domain model is then used to replace * as the input domain to SATPLAN 2006 solver [7], in order to find a solution to the planning problem presented in the section III. The plan built by the solver is the sequence of actions {unstack C A, putdown C, pickup B, stack B A, pickup C, stack C B, pickup D, stack D C}, which is indeed a solution to the same planning problem on the complete domain. VI. CONCLUSIONS This paper presents DIR action model learning method for incomplete STRIPS domain problems. The experiment
6 reported shows one case in which the algorithm starts with a 31% incomplete domain model (8 missing literals in a total of 26) and yields an inflated model with 7,6% of exceeding literals (28 over 26 literals). These exceeding literals remained only in one of the four operators, what guarantees that plans which do not apply the operator unstack (?x?y) are actual solutions in the complete model. Furthermore, only in cases when this operator is required to apply on a state such that block y is over some other block the solver will not yield a correct solution. Another remarkable aspect is that a low number of plan cases was sufficient to provide a reasonable improvement of complete domain s knowledge. In addition, correct plans generated contribute to compose the set of successful cases that may be reused for further applications. However, the results of the present research are too limited to provide means to assess accuracy and robustness of the proposed technique. DIR algorithm is at present under computational implementation. Future work consists on tests with more complex domain, models with different degrees of incompleteness and sets of plan traces with different sizes. It is also needed to assess the scalability of solutions and to compare performances to state-of-art action model learning algorithms. ACKNOWLEDGMENT The authors would like to thank Prof. Dr. Fabiano Silva, Prof. Dr. Luis Allan Künzle and Prof. Marcos Alexandre Castilho from DINF/UFPR for the support on this research and CAPES for the financial subvention. REFERENCES [1] P. Bertoli, M. Pistore, and P. Traverso, Automated composition of web services via planning in asynchronous domains, Artificial Intelligence Journal, vol. 174, pp , [2] J. Blythe, E. Deelman, and Y. Gil, Automatically composed workflows for grid environments, IEEE Intelligent Systems, vol. 19, pp , [3] S. N. Cresswell, T. L. McCluskey, and M. M. West, Acquisition of object-centered domain models from planning examples, in Proceedings of ICAPS 19, pp , [4] M. Ghallab, D. Nau, and P. Traverso, Automated planning: theory and practice, Morgan Kaufman: San Francisco, 2004, pp [5] J. Hoffmann, P. Bertoli, and M. Pistore, Web service compositions as planning, revisited: in between background theories and initial state uncertainty, in Proceedings of AAAI Conference on Artificial Intelligence 22, pp , [6] S. Kambhampati, Model-lite planning for the web age masses: the challenges of planning with incomplete and evolving domain models, in Proceedings of AAAI Conference on Artificial Intelligence 22, pp , [7] H. Kautz, B. Selman, and J. Hoffman, Satplan: planning as satisfiability, in Abstracts of International Planning Competition 5, [8] T. A. Nguyen, S. Kambhampati, and M. B. Do, Assessing and generating robust plans with partial domain models, in ICAPS Workshop on Planning under Uncertainty, [9] T. J. Walsh, and M. L. Littman, Efficient learning of action schemas and web-service descriptions, in Proceeding of AAAI Conference on Artificial Intelligence 23, pp , [10] C. Weber, and D. Bryce, Planning and acting in incomplete domains, in Proceedings of ICAPS 21, pp , [11] Q. Yang, K. Wu, and Y. Jiang, Learning action models from plan examples using weighted MAX-SAT, Artificial Intelligence Journal, vol. 171, pp , [12] L. S. Zettlemoyer, H. M. Pasula, and L. P. Kaelbling, Learning planning rules in noisy stochastic worlds, in Proceedings of AAAI Conference on Artificial Intelligence 20, pp , [13] H. H. Zhuo, T. Nguyen, and S. Kambhampati, Model-lite case-based planning, in Proceedings of AAAI Conference on Artificial intelligence 27, pp , [14] H. H. Zhuo, T. Nguyen, and S. Kambhampati, Refining incomplete planning domain models through plan traces, in Proceedings of IJCAI 23, pp , [15] H. H. Zhuo, Q. Yang, D. H. Hu, and L. Li, Learning complex action models with quantifiers and logical implications, Artificial Intelligence Journal, vol. 174, pp , 2010.
Model-Lite Case-Based Planning
Proceedings of the Twenty-Seventh AAAI Conference on Artificial Intelligence Model-Lite Case-Based Planning Hankz Hankui Zhuo a, Tuan Nguyen b, and Subbarao Kambhampati b a Dept. of Computer Science, Sun
More informationAction-Model Acquisition from Noisy Plan Traces
Action-Model Acquisition from Noisy Plan Traces Hankz Hankui Zhuo and Subbarao Kambhampati Dept. of Computer Science, Sun Yat-sen University, Guangzhou, China zhuohank@mail.sysu.edu.cn Dept. of Computer
More informationPlanning as Search. Progression. Partial-Order causal link: UCPOP. Node. World State. Partial Plans World States. Regress Action.
Planning as Search State Space Plan Space Algorihtm Progression Regression Partial-Order causal link: UCPOP Node World State Set of Partial Plans World States Edge Apply Action If prec satisfied, Add adds,
More informationOn Reduct Construction Algorithms
1 On Reduct Construction Algorithms Yiyu Yao 1, Yan Zhao 1 and Jue Wang 2 1 Department of Computer Science, University of Regina Regina, Saskatchewan, Canada S4S 0A2 {yyao, yanzhao}@cs.uregina.ca 2 Laboratory
More informationScienceDirect. Plan Restructuring in Multi Agent Planning
Available online at www.sciencedirect.com ScienceDirect Procedia Computer Science 46 (2015 ) 396 401 International Conference on Information and Communication Technologies (ICICT 2014) Plan Restructuring
More informationThe STRIPS Subset of PDDL for the Learning Track of IPC-08
The STRIPS Subset of PDDL for the Learning Track of IPC-08 Alan Fern School of Electrical Engineering and Computer Science Oregon State University April 9, 2008 This document defines two subsets of PDDL
More information3 No-Wait Job Shops with Variable Processing Times
3 No-Wait Job Shops with Variable Processing Times In this chapter we assume that, on top of the classical no-wait job shop setting, we are given a set of processing times for each operation. We may select
More informationIntelligent Agents. State-Space Planning. Ute Schmid. Cognitive Systems, Applied Computer Science, Bamberg University. last change: 14.
Intelligent Agents State-Space Planning Ute Schmid Cognitive Systems, Applied Computer Science, Bamberg University last change: 14. April 2016 U. Schmid (CogSys) Intelligent Agents last change: 14. April
More informationGoal Recognition in Incomplete STRIPS Domain Models
Goal Recognition in Incomplete STRIPS Domain Models Ramon Fraga Pereira and Felipe Meneguzzi Pontifical Catholic University of Rio Grande do Sul (PUCRS), Brazil Postgraduate Programme in Computer Science,
More informationValidating Plans with Durative Actions via Integrating Boolean and Numerical Constraints
Validating Plans with Durative Actions via Integrating Boolean and Numerical Constraints Roman Barták Charles University in Prague, Faculty of Mathematics and Physics Institute for Theoretical Computer
More informationQualitative Multi-faults Diagnosis Based on Automated Planning II: Algorithm and Case Study
Qualitative Multi-faults Diagnosis Based on Automated Planning II: Algorithm and Case Study He-xuan Hu, Anne-lise Gehin, and Mireille Bayart Laboratoire d Automatique, Génie Informatique & Signal, UPRESA
More informationAn Improved Upper Bound for the Sum-free Subset Constant
1 2 3 47 6 23 11 Journal of Integer Sequences, Vol. 13 (2010), Article 10.8.3 An Improved Upper Bound for the Sum-free Subset Constant Mark Lewko Department of Mathematics University of Texas at Austin
More informationUnifying and extending hybrid tractable classes of CSPs
Journal of Experimental & Theoretical Artificial Intelligence Vol. 00, No. 00, Month-Month 200x, 1 16 Unifying and extending hybrid tractable classes of CSPs Wady Naanaa Faculty of sciences, University
More informationA Model of Machine Learning Based on User Preference of Attributes
1 A Model of Machine Learning Based on User Preference of Attributes Yiyu Yao 1, Yan Zhao 1, Jue Wang 2 and Suqing Han 2 1 Department of Computer Science, University of Regina, Regina, Saskatchewan, Canada
More informationA New Method to Index and Query Sets
A New Method to Index and Query Sets Jorg Hoffmann and Jana Koehler Institute for Computer Science Albert Ludwigs University Am Flughafen 17 79110 Freiburg, Germany hoffmann koehler@informatik.uni-freiburg.de
More informationGenerating Macro-operators by Exploiting Inner Entanglements
Generating Macro-operators by Exploiting Inner Entanglements Lukáš Chrpa and Mauro Vallati and Thomas Leo McCluskey and Diane Kitchin PARK Research Group School of Computing and Engineering University
More informationOn the Relationships between Zero Forcing Numbers and Certain Graph Coverings
On the Relationships between Zero Forcing Numbers and Certain Graph Coverings Fatemeh Alinaghipour Taklimi, Shaun Fallat 1,, Karen Meagher 2 Department of Mathematics and Statistics, University of Regina,
More informationSet 9: Planning Classical Planning Systems. ICS 271 Fall 2013
Set 9: Planning Classical Planning Systems ICS 271 Fall 2013 Outline: Planning Classical Planning: Situation calculus PDDL: Planning domain definition language STRIPS Planning Planning graphs Readings:
More informationHandout 9: Imperative Programs and State
06-02552 Princ. of Progr. Languages (and Extended ) The University of Birmingham Spring Semester 2016-17 School of Computer Science c Uday Reddy2016-17 Handout 9: Imperative Programs and State Imperative
More information2386 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 52, NO. 6, JUNE 2006
2386 IEEE TRANSACTIONS ON INFORMATION THEORY, VOL. 52, NO. 6, JUNE 2006 The Encoding Complexity of Network Coding Michael Langberg, Member, IEEE, Alexander Sprintson, Member, IEEE, and Jehoshua Bruck,
More informationVS 3 : SMT Solvers for Program Verification
VS 3 : SMT Solvers for Program Verification Saurabh Srivastava 1,, Sumit Gulwani 2, and Jeffrey S. Foster 1 1 University of Maryland, College Park, {saurabhs,jfoster}@cs.umd.edu 2 Microsoft Research, Redmond,
More informationSTABILITY AND PARADOX IN ALGORITHMIC LOGIC
STABILITY AND PARADOX IN ALGORITHMIC LOGIC WAYNE AITKEN, JEFFREY A. BARRETT Abstract. Algorithmic logic is the logic of basic statements concerning algorithms and the algorithmic rules of deduction between
More informationEMPLOYING DOMAIN KNOWLEDGE TO IMPROVE AI PLANNING EFFICIENCY *
Iranian Journal of Science & Technology, Transaction B, Engineering, Vol. 29, No. B1 Printed in The Islamic Republic of Iran, 2005 Shiraz University EMPLOYING DOMAIN KNOWLEDGE TO IMPROVE AI PLANNING EFFICIENCY
More informationInternational Journal of Software and Web Sciences (IJSWS)
International Association of Scientific Innovation and Research (IASIR) (An Association Unifying the Sciences, Engineering, and Applied Research) ISSN (Print): 2279-0063 ISSN (Online): 2279-0071 International
More informationThe Encoding Complexity of Network Coding
The Encoding Complexity of Network Coding Michael Langberg Alexander Sprintson Jehoshua Bruck California Institute of Technology Email: mikel,spalex,bruck @caltech.edu Abstract In the multicast network
More informationChapter 15 Introduction to Linear Programming
Chapter 15 Introduction to Linear Programming An Introduction to Optimization Spring, 2015 Wei-Ta Chu 1 Brief History of Linear Programming The goal of linear programming is to determine the values of
More informationKing s Research Portal
King s Research Portal Link to publication record in King's Research Portal Citation for published version (APA): Krivic, S., Cashmore, M., Ridder, B. C., & Piater, J. (2016). Initial State Prediction
More informationFundamental Properties of Graphs
Chapter three In many real-life situations we need to know how robust a graph that represents a certain network is, how edges or vertices can be removed without completely destroying the overall connectivity,
More information3.4 Deduction and Evaluation: Tools Conditional-Equational Logic
3.4 Deduction and Evaluation: Tools 3.4.1 Conditional-Equational Logic The general definition of a formal specification from above was based on the existence of a precisely defined semantics for the syntax
More informationAn algorithm for Performance Analysis of Single-Source Acyclic graphs
An algorithm for Performance Analysis of Single-Source Acyclic graphs Gabriele Mencagli September 26, 2011 In this document we face with the problem of exploiting the performance analysis of acyclic graphs
More informationOn Constraint Problems with Incomplete or Erroneous Data
On Constraint Problems with Incomplete or Erroneous Data Neil Yorke-Smith and Carmen Gervet IC Parc, Imperial College, London, SW7 2AZ, U.K. nys,cg6 @icparc.ic.ac.uk Abstract. Real-world constraint problems
More informationAdvanced Algorithms Class Notes for Monday, October 23, 2012 Min Ye, Mingfu Shao, and Bernard Moret
Advanced Algorithms Class Notes for Monday, October 23, 2012 Min Ye, Mingfu Shao, and Bernard Moret Greedy Algorithms (continued) The best known application where the greedy algorithm is optimal is surely
More informationROUGH MEMBERSHIP FUNCTIONS: A TOOL FOR REASONING WITH UNCERTAINTY
ALGEBRAIC METHODS IN LOGIC AND IN COMPUTER SCIENCE BANACH CENTER PUBLICATIONS, VOLUME 28 INSTITUTE OF MATHEMATICS POLISH ACADEMY OF SCIENCES WARSZAWA 1993 ROUGH MEMBERSHIP FUNCTIONS: A TOOL FOR REASONING
More informationLearning a Negative Precondition
From: Proceedings of the Twelfth International FLAIRS Conference. Copyright 1999, AAAI (www.aaai.org). All rights reserved. Learning Opposite concept for Machine Planning Kang Soo Tae Department of Computer
More informationTowards Incremental Grounding in Tuffy
Towards Incremental Grounding in Tuffy Wentao Wu, Junming Sui, Ye Liu University of Wisconsin-Madison ABSTRACT Markov Logic Networks (MLN) have become a powerful framework in logical and statistical modeling.
More informationSatisfiability. Michail G. Lagoudakis. Department of Computer Science Duke University Durham, NC SATISFIABILITY
Satisfiability Michail G. Lagoudakis Department of Computer Science Duke University Durham, NC 27708 COMPSCI 271 - Spring 2001 DUKE UNIVERSITY Page 1 Why SAT? Historical Reasons The first NP-COMPLETE problem
More informationA proof-producing CSP solver: A proof supplement
A proof-producing CSP solver: A proof supplement Report IE/IS-2010-02 Michael Veksler Ofer Strichman mveksler@tx.technion.ac.il ofers@ie.technion.ac.il Technion Institute of Technology April 12, 2010 Abstract
More informationRelational Database: The Relational Data Model; Operations on Database Relations
Relational Database: The Relational Data Model; Operations on Database Relations Greg Plaxton Theory in Programming Practice, Spring 2005 Department of Computer Science University of Texas at Austin Overview
More informationArtificial Intelligence
Artificial Intelligence Other models of interactive domains Marc Toussaint University of Stuttgart Winter 2018/19 Basic Taxonomy of domain models Other models of interactive domains Basic Taxonomy of domain
More informationAlloy: A Lightweight Object Modelling Notation
Alloy: A Lightweight Object Modelling Notation Daniel Jackson, ACM Transactions on Software Engineering, 2002 Presented By: Steven Stewart, 2012-January-23 1 Alloy: 2002 to present Software is built on
More informationAutomated Service Composition using Heuristic Search
Automated Service Composition using Heuristic Search Harald Meyer and Mathias Weske Hasso-Plattner-Institute for IT-Systems-Engineering at the University of Potsdam Prof.-Dr.-Helmert-Strasse 2-3, 14482
More informationJoint Entity Resolution
Joint Entity Resolution Steven Euijong Whang, Hector Garcia-Molina Computer Science Department, Stanford University 353 Serra Mall, Stanford, CA 94305, USA {swhang, hector}@cs.stanford.edu No Institute
More informationWeb Service Usage Mining: Mining For Executable Sequences
7th WSEAS International Conference on APPLIED COMPUTER SCIENCE, Venice, Italy, November 21-23, 2007 266 Web Service Usage Mining: Mining For Executable Sequences MOHSEN JAFARI ASBAGH, HASSAN ABOLHASSANI
More informationDISCRETE-event dynamic systems (DEDS) are dynamic
IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY, VOL. 7, NO. 2, MARCH 1999 175 The Supervised Control of Discrete-Event Dynamic Systems François Charbonnier, Hassane Alla, and René David Abstract The supervisory
More informationA CSP Search Algorithm with Reduced Branching Factor
A CSP Search Algorithm with Reduced Branching Factor Igor Razgon and Amnon Meisels Department of Computer Science, Ben-Gurion University of the Negev, Beer-Sheva, 84-105, Israel {irazgon,am}@cs.bgu.ac.il
More informationLOGIC AND DISCRETE MATHEMATICS
LOGIC AND DISCRETE MATHEMATICS A Computer Science Perspective WINFRIED KARL GRASSMANN Department of Computer Science University of Saskatchewan JEAN-PAUL TREMBLAY Department of Computer Science University
More informationFast and Informed Action Selection for Planning with Sensing
Fast and Informed Action Selection for Planning with Sensing Alexandre Albore 1, Héctor Palacios 1, and Hector Geffner 2 1 Universitat Pompeu Fabra Passeig de Circumvalació 8 08003 Barcelona Spain 2 ICREA
More informationLearning Techniques for Pseudo-Boolean Solving and Optimization
Learning Techniques for Pseudo-Boolean Solving and Optimization José Faustino Fragoso Fremenin dos Santos September 29, 2008 Abstract The extension of conflict-based learning from Propositional Satisfiability
More information6. Lecture notes on matroid intersection
Massachusetts Institute of Technology 18.453: Combinatorial Optimization Michel X. Goemans May 2, 2017 6. Lecture notes on matroid intersection One nice feature about matroids is that a simple greedy algorithm
More informationPrimitive goal based ideas
Primitive goal based ideas Once you have the gold, your goal is to get back home s Holding( Gold, s) GoalLocation([1,1], s) How to work out actions to achieve the goal? Inference: Lots more axioms. Explodes.
More informationWhat is On / Off Policy?
What is On / Off Policy? Q learns how to perform optimally even when we are following a non-optimal policy In greedy, leaves no trace in Q SARSA is on-policy Learns the best policy given our systematic
More informationNDL A Modeling Language for Planning
NDL A Modeling Language for Planning Jussi Rintanen Aalto University Department of Computer Science February 20, 2017 Abstract This document describes the NDL modeling language for deterministic full information
More informationSystem Assistance in Structured Domain Model Development*
System Assistance in Structured Domain Model Development* Susanne Biundo and Werner Stephan German Research Center for Artificial Intelligence (DFKI) Stuhlsatzenhausweg 3 D-66123 Saarbriicken, Germany
More informationPartially Observable Markov Decision Processes. Silvia Cruciani João Carvalho
Partially Observable Markov Decision Processes Silvia Cruciani João Carvalho MDP A reminder: is a set of states is a set of actions is the state transition function. is the probability of ending in state
More informationCS 44 Exam #2 February 14, 2001
CS 44 Exam #2 February 14, 2001 Name Time Started: Time Finished: Each question is equally weighted. You may omit two questions, but you must answer #8, and you can only omit one of #6 or #7. Circle the
More informationA Short Introduction to Formal Specifications
A Short Introduction to Formal Specifications Prof. Dr. Hans J. Schneider Lehrstuhl für Programmiersprachen und Programmiermethodik Friedrich-Alexander-Universität Erlangen-Nürnberg 30.11.2000 Methods
More informationList of figures List of tables Acknowledgements
List of figures List of tables Acknowledgements page xii xiv xvi Introduction 1 Set-theoretic approaches in the social sciences 1 Qualitative as a set-theoretic approach and technique 8 Variants of QCA
More informationAND-OR GRAPHS APPLIED TO RUE RESOLUTION
AND-OR GRAPHS APPLIED TO RUE RESOLUTION Vincent J. Digricoli Dept. of Computer Science Fordham University Bronx, New York 104-58 James J, Lu, V. S. Subrahmanian Dept. of Computer Science Syracuse University-
More informationMathematical Logic Prof. Arindama Singh Department of Mathematics Indian Institute of Technology, Madras. Lecture - 37 Resolution Rules
Mathematical Logic Prof. Arindama Singh Department of Mathematics Indian Institute of Technology, Madras Lecture - 37 Resolution Rules If some literals can be unified, the same algorithm should be able
More informationA Correctness Proof for a Practical Byzantine-Fault-Tolerant Replication Algorithm
Appears as Technical Memo MIT/LCS/TM-590, MIT Laboratory for Computer Science, June 1999 A Correctness Proof for a Practical Byzantine-Fault-Tolerant Replication Algorithm Miguel Castro and Barbara Liskov
More informationA Review on Unsupervised Record Deduplication in Data Warehouse Environment
A Review on Unsupervised Record Deduplication in Data Warehouse Environment Keshav Unde, Dhiraj Wadhawa, Bhushan Badave, Aditya Shete, Vaishali Wangikar School of Computer Engineering, MIT Academy of Engineering,
More informationarxiv:submit/ [math.co] 9 May 2011
arxiv:submit/0243374 [math.co] 9 May 2011 Connectivity and tree structure in finite graphs J. Carmesin R. Diestel F. Hundertmark M. Stein 6 May, 2011 Abstract We prove that, for every integer k 0, every
More informationOrdering Problem Subgoals
Ordering Problem Subgoals Jie Cheng and Keki B. Irani Artificial Intelligence Laboratory Department of Electrical Engineering and Computer Science The University of Michigan, Ann Arbor, MI 48109-2122,
More informationEfficient Policy Analysis for Evolving Administrative Role Based Access Control 1
Efficient Policy Analysis for Evolving Administrative Role Based Access Control 1 Mikhail I. Gofman 1 and Ping Yang 2 1 Dept. of Computer Science, California State University of Fullerton, CA, 92834 2
More informationDipartimento di Elettronica Informazione e Bioingegneria. Cognitive Robotics. SATplan. Act1. Pre1. Fact. G. Gini Act2
Dipartimento di Elettronica Informazione e Bioingegneria Cognitive Robotics SATplan Pre1 Pre2 @ 2015 Act1 Act2 Fact why SAT (satisfability)? 2 Classical planning has been observed as a form of logical
More informationConstraint Satisfaction Problems
Constraint Satisfaction Problems CE417: Introduction to Artificial Intelligence Sharif University of Technology Spring 2013 Soleymani Course material: Artificial Intelligence: A Modern Approach, 3 rd Edition,
More informationAn Efficient Algorithm for Computing Non-overlapping Inversion and Transposition Distance
An Efficient Algorithm for Computing Non-overlapping Inversion and Transposition Distance Toan Thang Ta, Cheng-Yao Lin and Chin Lung Lu Department of Computer Science National Tsing Hua University, Hsinchu
More informationA Comparative Study of Data Mining Process Models (KDD, CRISP-DM and SEMMA)
International Journal of Innovation and Scientific Research ISSN 2351-8014 Vol. 12 No. 1 Nov. 2014, pp. 217-222 2014 Innovative Space of Scientific Research Journals http://www.ijisr.issr-journals.org/
More informationReformulating Constraint Models for Classical Planning
Reformulating Constraint Models for Classical Planning Roman Barták, Daniel Toropila Charles University, Faculty of Mathematics and Physics Malostranské nám. 2/25, 118 00 Praha 1, Czech Republic roman.bartak@mff.cuni.cz,
More informationOntology Merging: on the confluence between theoretical and pragmatic approaches
Ontology Merging: on the confluence between theoretical and pragmatic approaches Raphael Cóbe, Renata Wassermann, Fabio Kon 1 Department of Computer Science University of São Paulo (IME-USP) {rmcobe,renata,fabio.kon}@ime.usp.br
More informationA Fast Arc Consistency Algorithm for n-ary Constraints
A Fast Arc Consistency Algorithm for n-ary Constraints Olivier Lhomme 1 and Jean-Charles Régin 2 1 ILOG, 1681, route des Dolines, 06560 Valbonne, FRANCE 2 Computing and Information Science, Cornell University,
More informationSmall Formulas for Large Programs: On-line Constraint Simplification In Scalable Static Analysis
Small Formulas for Large Programs: On-line Constraint Simplification In Scalable Static Analysis Isil Dillig, Thomas Dillig, Alex Aiken Stanford University Scalability and Formula Size Many program analysis
More informationRESULTS ON TRANSLATING DEFAULTS TO CIRCUMSCRIPTION. Tomasz Imielinski. Computer Science Department Rutgers University New Brunswick, N.
RESULTS ON TRANSLATING DEFAULTS TO CIRCUMSCRIPTION Tomasz Imielinski Computer Science Department Rutgers University New Brunswick, N.J 08905 ABSTRACT In this paper we define different concepts, of translating
More informationCSE Theory of Computing Fall 2017 Project 1-SAT Solving
CSE 30151 Theory of Computing Fall 2017 Project 1-SAT Solving Version 3: Sept. 21, 2017 The purpose of this project is to gain an understanding of one of the most central problems of computing: Boolean
More informationCHAPTER 9. GRAPHS 310. Figure 9.1 An undirected graph. Figure 9.2 A directed graph
Chapter 9 Graphs Often we need to model relationships that can be expressed using a set of pairs. Examples include distances between points on a map, links in a communications network, precedence constraints
More informationA Case-Based Approach to Heuristic Planning
Noname manuscript No. (will be inserted by the editor) A Case-Based Approach to Heuristic Planning Tomás de la Rosa Angel García-Olaya Daniel Borrajo the date of receipt and acceptance should be inserted
More informationFast algorithms for max independent set
Fast algorithms for max independent set N. Bourgeois 1 B. Escoffier 1 V. Th. Paschos 1 J.M.M. van Rooij 2 1 LAMSADE, CNRS and Université Paris-Dauphine, France {bourgeois,escoffier,paschos}@lamsade.dauphine.fr
More informationValue Added Association Rules
Value Added Association Rules T.Y. Lin San Jose State University drlin@sjsu.edu Glossary Association Rule Mining A Association Rule Mining is an exploratory learning task to discover some hidden, dependency
More information3.4 Data-Centric workflow
3.4 Data-Centric workflow One of the most important activities in a S-DWH environment is represented by data integration of different and heterogeneous sources. The process of extract, transform, and load
More informationCSE 215: Foundations of Computer Science Recitation Exercises Set #9 Stony Brook University. Name: ID#: Section #: Score: / 4
CSE 215: Foundations of Computer Science Recitation Exercises Set #9 Stony Brook University Name: ID#: Section #: Score: / 4 Unit 14: Set Theory: Definitions and Properties 1. Let C = {n Z n = 6r 5 for
More informationVisibilty: Finding the Staircase Kernel in Orthogonal Polygons
American Journal of Computational and Applied Mathematics 2012, 2(2): 17-24 DOI: 10.5923/j.ajcam.20120202.04 Visibilty: Finding the Staircase Kernel in Orthogonal Polygons Stefan A. Pape, Tzvetalin S.
More informationOntology based Model and Procedure Creation for Topic Analysis in Chinese Language
Ontology based Model and Procedure Creation for Topic Analysis in Chinese Language Dong Han and Kilian Stoffel Information Management Institute, University of Neuchâtel Pierre-à-Mazel 7, CH-2000 Neuchâtel,
More informationUsing Coarse State Space Abstractions to Detect Mutex Pairs
Using Coarse State Space Abstractions to Detect Mutex Pairs Mehdi Sadeqi Computer Science Department University of Regina Regina, SK, Canada S4S 0A2 (sadeqi2m@cs.uregina.ca) Robert C. Holte Computing Science
More informationTest Cases Generation from UML Activity Diagrams
Eighth ACIS International Conference on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Computing Test Cases Generation from UML Activity Diagrams Hyungchoul Kim, Sungwon
More informationOn The Theoretical Foundation for Data Flow Analysis in Workflow Management
Association for Information Systems AIS Electronic Library (AISeL) AMCIS 2005 Proceedings Americas Conference on Information Systems (AMCIS) 2005 On The Theoretical Foundation for Data Flow Analysis in
More informationAn Improved Separation of Regular Resolution from Pool Resolution and Clause Learning
An Improved Separation of Regular Resolution from Pool Resolution and Clause Learning Maria Luisa Bonet and Sam Buss Theory and Applications of Satisfiability Testing SAT 2012 Trento, Italy July 17, 2012
More informationConstraint Satisfaction Problems
Constraint Satisfaction Problems Tuomas Sandholm Carnegie Mellon University Computer Science Department [Read Chapter 6 of Russell & Norvig] Constraint satisfaction problems (CSPs) Standard search problem:
More informationScan Scheduling Specification and Analysis
Scan Scheduling Specification and Analysis Bruno Dutertre System Design Laboratory SRI International Menlo Park, CA 94025 May 24, 2000 This work was partially funded by DARPA/AFRL under BAE System subcontract
More informationEfficient Circuit to CNF Conversion
Efficient Circuit to CNF Conversion Panagiotis Manolios and Daron Vroon College of Computing, Georgia Institute of Technology, Atlanta, GA, 30332, USA http://www.cc.gatech.edu/home/{manolios,vroon} Abstract.
More informationOn Computing Minimum Size Prime Implicants
On Computing Minimum Size Prime Implicants João P. Marques Silva Cadence European Laboratories / IST-INESC Lisbon, Portugal jpms@inesc.pt Abstract In this paper we describe a new model and algorithm for
More informationComputer Science Technical Report
Computer Science Technical Report Feasibility of Stepwise Addition of Multitolerance to High Atomicity Programs Ali Ebnenasir and Sandeep S. Kulkarni Michigan Technological University Computer Science
More informationLecture : Topological Space
Example of Lecture : Dr. Department of Mathematics Lovely Professional University Punjab, India October 18, 2014 Outline Example of 1 2 3 Example of 4 5 6 Example of I Topological spaces and continuous
More informationICS 606. Intelligent Autonomous Agents 1
Intelligent utonomous gents ICS 606 / EE 606 Fall 2011 Nancy E. Reed nreed@hawaii.edu Lecture #4 Practical Reasoning gents Intentions Planning Means-ends reasoning The blocks world References Wooldridge
More informationA graph is finite if its vertex set and edge set are finite. We call a graph with just one vertex trivial and all other graphs nontrivial.
2301-670 Graph theory 1.1 What is a graph? 1 st semester 2550 1 1.1. What is a graph? 1.1.2. Definition. A graph G is a triple (V(G), E(G), ψ G ) consisting of V(G) of vertices, a set E(G), disjoint from
More information6 th International Planning Competition: Uncertainty Part
6 th International Planning Competition: Uncertainty Part Daniel Bryce SRI International bryce@ai.sri.com Olivier Buffet LORIA-INRIA olivier.buffet@loria.fr Abstract The 6 th International Planning Competition
More informationTemporal Exception Prediction for Loops in Resource Constrained Concurrent Workflows
emporal Exception Prediction for Loops in Resource Constrained Concurrent Workflows Iok-Fai Leong, Yain-Whar Si Faculty of Science and echnology, University of Macau {henryifl, fstasp}@umac.mo Abstract
More informationEvidence for Invariants in Local Search
This paper appears in the Proceedings of the Fourteenth National Conference on Artificial Intelligence (AAAI-97), Providence, RI, 1997. Copyright 1997 American Association for Artificial Intelligence.
More informationTopology Homework 3. Section Section 3.3. Samuel Otten
Topology Homework 3 Section 3.1 - Section 3.3 Samuel Otten 3.1 (1) Proposition. The intersection of finitely many open sets is open and the union of finitely many closed sets is closed. Proof. Note that
More informationDeductive Methods, Bounded Model Checking
Deductive Methods, Bounded Model Checking http://d3s.mff.cuni.cz Pavel Parízek CHARLES UNIVERSITY IN PRAGUE faculty of mathematics and physics Deductive methods Pavel Parízek Deductive Methods, Bounded
More informationELEMENTARY NUMBER THEORY AND METHODS OF PROOF
CHAPTER 4 ELEMENTARY NUMBER THEORY AND METHODS OF PROOF Copyright Cengage Learning. All rights reserved. SECTION 4.6 Indirect Argument: Contradiction and Contraposition Copyright Cengage Learning. All
More information