Improving the performance of local real-time scheduling

Abstract. Mining association rules is an important task, even though the number of rules discovered is often huge. A possible solution to this problem, is to use the Formal Concept Analysis (FCA) mathematical settings to restrict rules extraction to a generic basis of association rules. This one is considered as a reduced set to which we can apply appropriate inference mechanisms to derive redundant rules. In this paper, we introduce a new minimal generic basis MGB of non-redundant association rules based on the augmented Iceberg Galois lattice. The proposed approach involves the inference mechanisms used and a set of experiments applied to several real and synthetic databases. Carried out experiments showed important benefits in terms of reduction in the number of generic rules extracted. We present also a new framework for generating and visually exploring the minimal generic basis MGB. 1

concepts are relevant when transferring ligand-receptor interactions to a computational level are discussed. Various approaches are feasible and methods can be classified according to which approach they take. As an example of a docking method the tool FlexX is described in detail. Similar concepts

Access to remote data is one of the principal challenges of grid computing. While performing I/O, grid applications must be prepared for server crashes, performance variations, and exhausted resources. To achieve high throughput in such a hostile environment, applications need a resilient service

Weanalyzetheconceptofmalignness,whichisthe propertyofprobabilityensemblesofmakingtheaver-agecaserunningtimeequaltotheworstcaserunning timeforaclassofalgorithms.Wederivelowerand upperboundsonthecomplexityofmalignensembles, whicharetightforexponentialtimealgorithms,and whichshowthatnopolynomialtimecomputablema-lignensembleexistsfortheclassofpolynomialtime algorithms.Furthermore,weshowthatfornoclassof superlinearalgorithmsapolynomialtimecomputable malignensembleexists,unlesseverylanguageinPhas anexpectedpolynomialtimeconstructor. 1Introduction Theaveragecasetimecomplexityofspecicalgorithms hasforanumberofyearsbeenanactiveareaofre-search,oftenshowingsignicantimprovementoverthe worstcasecomplexitywhenspecicdistributionsofthe inputswereassumed.Recently,LiandVitanyi[4]stud-iedtheSolomono-Levinmeasuremandfoundthat whentheinputstoanyalgorithmaredistributedac-cordingtothismeasure,itholdsthatthealgorithm's averagecasecomplexityisofthesameorderofmagni-tudeasitsworstcasecomplexity.Moreprecisely,for somec>0 X x2nm(x)

ualifying Project (MQP) is required of all undergraduate students at Worcester Polytechnic Institute. The MQP within the Computer Science Department is a capstone experience, requiring one unit of work, that gives students practice on applying the fundamentals and skills they have learned to a

Thisresearchprojectaimstousemachinelearningtechniquestoimprovetheperformanceofthreedimensionalvisionsystems.Buildingonourearlier work,ourapproachrepresentsandorganizesmodelsof objectclassesinahierarchyofprobabilisticconcepts, anditusesBayesianinferencemethodstofocusattention,recognizeobjectsinimages,andmakepredictions aboutoccludedparts.Thelearningprocessinvolvesnot onlyupdatingoftheprobabilisticdescriptionsintheconcepthierarchybutalsoinvolveschangesinthestructure ofmemory,includingthecreationofnovelcategories,the mergingofsimilarclasses,andtheeliminationofunnecessaryones.Anevaluationmetricbasedonprobability theoryguidesdecisionsaboutsuchstructuralchanges, andbackgroundknowledgeaboutfunctionandgeneric objectclassesfurtherconstrainsthelearningprocess. Weplantocarryoutsystematicexperimentstodeterminetheabilityofthisapproachtoimprovebothclassicationaccuracyandpredictiveabilityonnovelimages.

Inthispaperwereportonanapproachtolearningobject modelsforuseinrecognitionandreconstruction.Our frameworkrepresentsobjectsinanimageusinggeneralizedcylindersandorganizesknowledgeaboutclasses ofobjectsinaBayesiannetwork.Therecognitionprocessinvolvespropagatingevidencethroughthisinference network,whereaslearningreliesonupdatingofthenetwork'sconditionalprobabilitiesbasedontrainingcases. Wereportpreliminaryexperimentalresultswithsyntheticdatathatsuggestourmethodimprovesitsrecognitionaccuracywithexperience.Wealsoconsiderour framework'srelationtootherresearchonlearningobject knowledgeforimageunderstanding. Theimage-understandingprocessreliesonaccurate knowledge.Thisstatementholdsforalllevelsofvisualprocessing,butseemsespeciallytrueatthelater stages,whereobjectrecognitionandreconstructionrequiremodelsofobjectsorobjectclassesthatoccurinthe domain.Andthoughonecould,inprinciple,entersuch

Abstract not found

A code sandwich consists of before, after, and body code. Typically, the before code makes some change, the body code works in the context of the change, and the after code undoes the change. A code sandwich must guarantee that its after code will execute if its before code has executed, even if exceptions arise. This pattern is common to many programming situations, and most modern languages have some language-level support for expressing it. We survey support for code sandwiches in several programming languages and proposed language extensions. We explain why such support can improve a program, consider related features, and discuss desirable properties that a language can provide its programmers. We relate these properties to Jyro, our code sandwich extension to Java. We examine two large, mature open-source programs, find numerous sandwich-related bugs and readability issues, and demonstrate how they might be avoided using our Jyro implementation. 1