Luc De Raedt – författare

An intelligent agent interacting with the real world will encounter individual people, courses, test results, drugs prescriptions, chairs, boxes, etc., and needs to reason about properties of these individuals and relations among them as well as cope with uncertainty. Uncertainty has been studied in probability theory and graphical models, and relations have been studied in logic, in particular in the predicate calculus and its extensions. This book examines the foundations of combining logic and probability into what are called relational probabilistic models. It introduces representations, inference, and learning techniques for probability, logic, and their combinations. The book focuses on two representations in detail: Markov logic networks, a relational extension of undirected graphical models and weighted first-order predicate calculus formula, and Problog, a probabilistic extension of logic programs that can also be viewed as a Turing-complete relational extension of Bayesian networks.

An intelligent agent interacting with the real world will encounter individual people, courses, test results, drugs prescriptions, chairs, boxes, etc., and needs to reason about properties of these individuals and relations among them as well as cope with uncertainty. Uncertainty has been studied in probability theory and graphical models, and relations have been studied in logic, in particular in the predicate calculus and its extensions. This book examines the foundations of combining logic and probability into what are called relational probabilistic models. It introduces representations, inference, and learning techniques for probability, logic, and their combinations. The book focuses on two representations in detail: Markov logic networks, a relational extension of undirected graphical models and weighted first-order predicate calculus formula, and Problog, a probabilistic extension of logic programs that can also be viewed as a Turing-complete relational extension of Bayesian networks.

A successful integration of constraint programming and data mining has the potential to lead to a new ICT paradigm with far reaching implications. It could change the face of data mining and machine learning, as well as constraint programming technology. It would not only allow one to use data mining techniques in constraint programming to identify and update constraints and optimization criteria, but also to employ constraints and criteria in data mining and machine learning in order to discover models compatible with prior knowledge. This book reports on some key results obtained on this integrated and cross- disciplinary approach within the European FP7 FET Open project no. 284715 on “Inductive Constraint Programming” and a number of associated workshops and Dagstuhl seminars. The book is structured in five parts: background; learning to model; learning to solve; constraint programming for data mining; and showcases.

Iusethetermlogicalandrelationallearning torefertothesub?eldofarti?cial intelligence,machinelearninganddataminingthatisconcernedwithlearning in expressive logical or relational representations. It is the union of inductive logic programming, (statistical) relational learning and multi-relational data mining, which all have contributed techniques for learning from data in re- tional form. Even though some early contributions to logical and relational learning are about forty years old now, it was only with the advent of - ductive logic programming in the early 1990s that the ?eld became popular. Whereas initial work was often concerned with logical (or logic programming) issues,thefocushasrapidlychangedtothediscoveryofnewandinterpretable knowledge from structured data, often in the form of rules, and soon imp- tant successes in applications in domains such as bio- and chemo-informatics and computational linguistics were realized. Today, the challenges and opp- tunities of dealing with structured data and knowledge have been taken up by the arti?cial intelligence community at large and form the motivation for a lot of ongoing research. Indeed, graph, network and multi-relational data mining are now popular themes in data mining, and statistical relational learning is receiving a lot of attention in the machine learning and uncertainty in art- cial intelligence communities. In addition, the range of tasks for which logical and relational techniques have been developed now covers almost all machine learning and data mining tasks.

The interconnected ideas of inductive databases and constraint-based mining are appealing and have the potential to radically change the theory and practice of data mining and knowledge discovery. This book reports on the results of the European IST project "cInQ" (consortium on knowledge discovery by Inductive Queries) and its final workshop entitled Constraint-Based Mining and Inductive Databases organized in Hinterzarten, Germany in March 2004.

This book constitutes the refereed proceedings of the 5th European Conference on Principles of Data Mining and Knowledge Discovery, PKDD 2001, held in Freiburg, Germany, in September 2001.The 40 revised full papers presented together with four invited contributions were carefully reviewed and selected from close to 100 submissions. Among the topics addressed are hidden Markov models, text summarization, supervised learning, unsupervised learning, demographic data analysis, phenotype data mining, spatio-temporal clustering, Web-usage analysis, association rules, clustering algorithms, time series analysis, rule discovery, text categorization, self-organizing maps, filtering, reinforcemant learning, support vector machines, visual data mining, and machine learning.

This book constitutes the refereed proceedings of the 12th European Conference on Machine Learning, ECML 2001, held in Freiburg, Germany, in September 2001.The 50 revised full papers presented together with four invited contributions were carefully reviewed and selected from a total of 140 submissions. Among the topics covered are classifier systems, naive-Bayes classification, rule learning, decision tree-based classification, Web mining, equation discovery, inductive logic programming, text categorization, agent learning, backpropagation, reinforcement learning, sequence prediction, sequential decisions, classification learning, sampling, and semi-supervised learning.

This book constitutes the refereed proceedings of the 5th European Conference on Principles of Data Mining and Knowledge Discovery, PKDD 2001, held in Freiburg, Germany, in September 2001.The 40 revised full papers presented together with four invited contributions were carefully reviewed and selected from close to 100 submissions. Among the topics addressed are hidden Markov models, text summarization, supervised learning, unsupervised learning, demographic data analysis, phenotype data mining, spatio-temporal clustering, Web-usage analysis, association rules, clustering algorithms, time series analysis, rule discovery, text categorization, self-organizing maps, filtering, reinforcemant learning, support vector machines, visual data mining, and machine learning.

This book constitutes the refereed proceedings of the 12th European Conference on Machine Learning, ECML 2001, held in Freiburg, Germany, in September 2001.The 50 revised full papers presented together with four invited contributions were carefully reviewed and selected from a total of 140 submissions. Among the topics covered are classifier systems, naive-Bayes classification, rule learning, decision tree-based classification, Web mining, equation discovery, inductive logic programming, text categorization, agent learning, backpropagation, reinforcement learning, sequence prediction, sequential decisions, classification learning, sampling, and semi-supervised learning.

This volume contains the proceedings of the European Conference on Machine Learning 1994, which continues the tradition of earlier meetings and which is a major forum for the presentation of the latest and most significant results in machine learning. Machine learning is one of the most important subfields of artificial intelligence and computer science, as it is concerned with the automation of learning processes. This volume contains two invited papers, 19 regular papers, and 25 short papers carefully reviewed and selected from in total 88 submissions. The papers describe techniques, algorithms, implementations, and experiments in the area of machine learning.

Iusethetermlogicalandrelationallearning torefertothesub?eldofarti?cial intelligence,machinelearninganddataminingthatisconcernedwithlearning in expressive logical or relational representations. It is the union of inductive logic programming, (statistical) relational learning and multi-relational data mining, which all have contributed techniques for learning from data in re- tional form. Even though some early contributions to logical and relational learning are about forty years old now, it was only with the advent of - ductive logic programming in the early 1990s that the ?eld became popular. Whereas initial work was often concerned with logical (or logic programming) issues,thefocushasrapidlychangedtothediscoveryofnewandinterpretable knowledge from structured data, often in the form of rules, and soon imp- tant successes in applications in domains such as bio- and chemo-informatics and computational linguistics were realized. Today, the challenges and opp- tunities of dealing with structured data and knowledge have been taken up by the arti?cial intelligence community at large and form the motivation for a lot of ongoing research. Indeed, graph, network and multi-relational data mining are now popular themes in data mining, and statistical relational learning is receiving a lot of attention in the machine learning and uncertainty in art- cial intelligence communities. In addition, the range of tasks for which logical and relational techniques have been developed now covers almost all machine learning and data mining tasks.

One of the key open questions within arti?cial intelligence is how to combine probability and logic with learning. This question is getting an increased - tentioninseveraldisciplinessuchasknowledgerepresentation,reasoningabout uncertainty, data mining, and machine learning simulateously, resulting in the newlyemergingsub?eldknownasstatisticalrelationallearningandprobabil- ticinductivelogicprogramming.Amajordriving forceisthe explosivegrowth in the amount of heterogeneous data that is being collected in the business and scienti?c world. Example domains include bioinformatics, chemoinform- ics, transportation systems, communication networks, social network analysis, linkanalysis,robotics,amongothers.Thestructuresencounteredcanbeass- pleassequencesandtrees(suchasthosearisinginproteinsecondarystructure predictionandnaturallanguageparsing)orascomplexascitationgraphs,the WorldWideWeb,andrelationaldatabases. This book providesan introduction to this ?eld with an emphasison those methods based on logic programming principles. The book is also the main resultofthesuccessfulEuropeanISTFETprojectno.FP6-508861onAppli- tionofProbabilisticInductiveLogicProgramming(APRILII,2004-2007).This projectwascoordinatedbytheAlbertLudwigsUniversityofFreiburg(Germany, Luc De Raedt) and the partners were Imperial College London (UK, Stephen MuggletonandMichaelSternberg),theHelsinkiInstituteofInformationTe- nology(Finland,HeikkiMannila),theUniversit' adegliStudidiFlorence(Italy, PaoloFrasconi),andtheInstitutNationaldeRechercheenInformatiqueet- tomatiqueRocquencourt(France,FrancoisFages).Itwasconcernedwiththeory, implementationsandapplicationsofprobabilisticinductivelogicprogramming. Thisstructureisalsore?ectedinthebook. The book starts with an introductory chapter to "Probabilistic Inductive LogicProgramming"byDeRaedtandKersting.Inasecondpart,itprovidesa detailedoverviewofthemostimportantprobabilisticlogiclearningformalisms and systems. We are very pleased and proud that the scientists behind the key probabilistic inductive logic programming systems (also those developed outside the APRIL project) have kindly contributed a chapter providing an overviewoftheircontributions.Thisincludes:relationalsequencelearningte- niques (Kersting et al.), using kernels with logical representations (Frasconi andPasserini),MarkovLogic(Domingosetal.), the PRISMsystem (Satoand Kameya),CLP(BN)(SantosCostaetal.),BayesianLogicPrograms(Kersting andDeRaedt),andtheIndependentChoiceLogic(Poole).Thethirdpartthen provides a detailed account of some show-caseapplications of probabilistic - ductive logic programming, more speci?cally: in protein fold discovery (Chen et al.), haplotyping (Landwehr and Mielik. ainen) and systems biology (Fages andSoliman). The ?nal parttouchesupon sometheoreticalinvestigationsand VI Preface includes chaptersonbehavioralcomparisonof probabilisticlogicprogramming representations(MuggletonandChen)andamodel-theoreticexpressivityan- ysis(Jaeger).

Iusethetermlogicalandrelationallearning torefertothesub?eldofarti?cial intelligence,machinelearninganddataminingthatisconcernedwithlearning in expressive logical or relational representations. It is the union of inductive logic programming, (statistical) relational learning and multi-relational data mining, which all have contributed techniques for learning from data in re- tional form. Even though some early contributions to logical and relational learning are about forty years old now, it was only with the advent of - ductive logic programming in the early 1990s that the ?eld became popular. Whereas initial work was often concerned with logical (or logic programming) issues,thefocushasrapidlychangedtothediscoveryofnewandinterpretable knowledge from structured data, often in the form of rules, and soon imp- tant successes in applications in domains such as bio- and chemo-informatics and computational linguistics were realized. Today, the challenges and opp- tunities of dealing with structured data and knowledge have been taken up by the arti?cial intelligence community at large and form the motivation for a lot of ongoing research. Indeed, graph, network and multi-relational data mining are now popular themes in data mining, and statistical relational learning is receiving a lot of attention in the machine learning and uncertainty in art- cial intelligence communities. In addition, the range of tasks for which logical and relational techniques have been developed now covers almost all machine learning and data mining tasks.