ebook img

Decision Trees with Hypotheses PDF

148 Pages·2022·2.024 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Decision Trees with Hypotheses

Synthesis Lectures on Intelligent Technologies Mohammad Azad · Igor Chikalov · Shahid Hussain · Mikhail Moshkov · Beata Zielosko Decision Trees with Hypotheses Synthesis Lectures on Intelligent Technologies SeriesEditor JanuszKacprzyk,SystemsResearchInstitute,PolishAcademyofScience,Warsaw, Poland Synthesis Lectures on Intelligent Technologies provides highly interdisciplinary research with the potential to change the fundamental principles of our society. It covers appli- cations such as Intelligent Transportation, Humanoids, Self-Driving Cars, IoT, Ambient Intelligence, Smart Cities, Human-computer Interaction, Computational Intelligence, Industry 4.0, Medical Robotics, or Data Science. Synthesis Lectures on Intelligent Tech- nologies brings together up-to-date resources from trusted authors working around the world in all aspects of Intelligent Systems. Mohammad Azad · Igor Chikalov · Shahid Hussain · Mikhail Moshkov · Beata Zielosko Decision Trees with Hypotheses MohammadAzad IgorChikalov DepartmentofComputerScience IntelCorporation CollegeofComputerandInformationSciences Chandler,AZ,USA JoufUniversity Sakakah,SaudiArabia MikhailMoshkov Computer,ElectricalandMathematical ShahidHussain SciencesandEngineeringDivision DepartmentofComputerScience ComputationalBioscienceResearchCenter SchoolofMathematicsandComputerScience KingAbdullahUniversityofScience InstituteofBusinessAdministration andTechnology Karachi,Pakistan Thuwal,SaudiArabia BeataZielosko FacultyofScienceandTechnology InstituteofComputerScience UniversityofSilesiainKatowice Sosnowiec,Poland ISSN2731-6912 ISSN2731-6920 (electronic) SynthesisLecturesonIntelligentTechnologies ISBN978-3-031-08584-0 ISBN978-3-031-08585-7 (eBook) https://doi.org/10.1007/978-3-031-08585-7 ©TheEditor(s)(ifapplicable)andTheAuthor(s),underexclusivelicensetoSpringerNatureSwitzerlandAG 2022 Thisworkissubjecttocopyright.AllrightsaresolelyandexclusivelylicensedbythePublisher,whetherthewhole orpartofthematerialisconcerned,specificallytherightsoftranslation,reprinting,reuseofillustrations,recitation, broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionorinformationstorage andretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodologynowknownor hereafterdeveloped. Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublicationdoes notimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevantprotective lawsandregulationsandthereforefreeforgeneraluse. Thepublisher,theauthors,andtheeditorsaresafetoassumethattheadviceandinformationinthisbookare believedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsortheeditorsgive awarranty,expressedorimplied,withrespecttothematerialcontainedhereinorforanyerrorsoromissionsthat mayhavebeenmade.Thepublisherremainsneutralwithregardtojurisdictionalclaimsinpublishedmapsand institutionalaffiliations. ThisSpringerimprintispublishedbytheregisteredcompanySpringerNatureSwitzerlandAG Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland Toourfamilies Preface Decision trees are widely used in many areas of computer science and related fields as classifiers, as a means for knowledge representation, and as algorithms to solve various problems.Theyarestudied,inparticular,intesttheory,roughsettheory,andexactlearn- ing. These theories are closely related: attributes from rough set theory and test theory correspondtomembershipqueriesfromexactlearning.Exactlearningstudiesadditionally the so-called equivalence queries. In this book, we added to the model considered in test theory and rough set theory the notion of a hypothesis that allowed us to use an analog of equivalence queries and studied decision trees using various combinations of attributes, hypotheses, and proper hypotheses (an analog of proper equivalence queries). The two main goals of this book are (i) to create tools for the experimental and the- oretical study of decision trees with hypotheses and (ii) to compare these decision trees with conventional decision trees that use only queries, each of which is based on one attribute. We got both experimental and theoretical results showing that the decision trees with hypotheses can have less complexity than the conventional decision trees. These results open up some prospects for using decision trees with hypotheses as a means for knowledge representation and as algorithms for computation of the Boolean functions. Theobtainedtheoreticalresultsandtoolsdesignedforthestudyofdecisiontreeswith hypothesescanbehelpfulforresearcherswhousedecisiontreesandrulesindataanalysis. This book can also be used for the creation of courses for graduate students. Sakaka, Saudi Arabia Mohammad Azad Chandler, USA Igor Chikalov Karachi, Pakistan Shahid Hussain Thuwal, Saudi Arabia Mikhail Moshkov Sosnowiec, Poland Beata Zielosko March 2022 vii viii Preface Acknowledgements WearegreatlyindebtedtoKingAbdullahUniversityofScienceandTechnol- ogyforitsimmensesupport. WearethankfultoProf.AndrzejSkowronandtoAlexanderKnyazevforstimulatingdiscussions. WeextendanexpressionofgratitudetoProf.JanuszKacprzyk,Dr.ThomasDitzinger,andthe SeriesSynthesisLecturesonIntelligentTechnologiesstaffatSpringerfortheirsupportinmaking thisbookpossible. Contents 1 Introduction ......................................................... 1 1.1 Part I. Decision Tables ............................................ 2 1.2 Part II. Infinite Binary Information Systems and Infinite Families of Concepts ..................................................... 4 1.3 Prospects of Using Decision Trees with Hypotheses .................. 5 1.4 Use of Book ..................................................... 7 References ........................................................... 8 PartI DecisionTables 2 MainNotions ........................................................ 13 2.1 Decision Tables and Uncertainty Measures .......................... 13 2.2 Decision Trees ................................................... 14 2.3 Decision Rules Derived from Decision Trees ........................ 16 References ........................................................... 17 3 DynamicProgrammingAlgorithmsforMinimizationofDecisionTree Complexity .......................................................... 19 3.1 Construction of Directed Acyclic Graph Δ(T) ....................... 20 3.2 Minimizing the Depth ............................................ 21 3.3 Minimizing the Number of Realizable Nodes ........................ 25 3.4 Minimizing the Number of Realizable Terminal Nodes ................ 28 3.5 Minimizing the Number of Working Nodes .......................... 30 3.6 On Number of Realizable Terminal Nodes .......................... 31 3.7 Results of Experiments ........................................... 34 3.7.1 Depth .................................................... 35 3.7.2 Number of Realizable Nodes ................................ 36 3.7.3 Number of Realizable Terminal Nodes ....................... 37 3.7.4 Number of Working Nodes ................................. 38 3.8 Conclusions ..................................................... 40 References ........................................................... 40 ix x Contents 4 ConstructionofOptimalDecisionTreesandDerivingDecisionRules fromThem .......................................................... 41 4.1 Construction of Decision Trees with Minimum Depth ................. 42 4.2 Construction of Decision Trees with Minimum Number of Working Nodes .......................................................... 45 4.3 On Construction of Optimal Decision Trees for L and L ............. 48 t 4.4 Results of Experiments ........................................... 49 4.4.1 Decision Trees with Minimum Depth ......................... 49 4.4.2 Decision Trees with Minimum Number of Working Nodes ...... 50 4.4.3 Analysis of Experimental Results ............................ 53 4.5 Conclusions ..................................................... 53 References ........................................................... 53 5 GreedyAlgorithmsforConstructionofDecisionTreeswithHypotheses ... 55 5.1 Greedy Algorithms ............................................... 56 5.2 Results of Experiments on Decision Tables from UCI ML Repository ... 57 5.2.1 Results for Misclassification Error me ........................ 57 5.2.2 Results for Relative Misclassification Errorrme ............... 59 5.2.3 Results for Entropy ent ..................................... 61 5.2.4 Results for Gini Index gini ................................. 63 5.2.5 Results for Uncertainty Measure R ........................... 65 5.3 Results of Experiments on Randomly Generated Boolean Functions .... 68 5.4 Analysis of Experimental Results .................................. 71 5.5 Conclusions ..................................................... 71 References ........................................................... 72 6 DecisionTreeswithHypothesesforRecognitionofMonotoneBoolean FunctionsandforSorting ............................................. 73 6.1 Problem of Recognition of Monotone Boolean Functions .............. 74 6.1.1 Basic Notions and Notation ................................. 74 6.1.2 Results of Experiments ..................................... 75 6.2 Problem of Sorting ............................................... 77 6.2.1 Basic Notions and Notation ................................. 77 6.2.2 Results of Experiments ..................................... 78 6.3 Conclusions ..................................................... 80 References ........................................................... 80 PartII BinaryInformationSystemsandInfiniteFamiliesofConcepts 7 InfiniteBinaryInformationSystems.DecisionTreesofTypes1,2,and3 ... 83 7.1 Basic Notions .................................................... 84 7.2 Main Results .................................................... 86 7.3 Proofs of Theorems 7.1 and 7.2 .................................... 88

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.