Inductive logic programming : techniques and applications



Inductive logic programming : techniques and applications

Nada Lavrač and Sašo Džeroski

(Ellis Horwood series in artificial intelligence)

E. Horwood, c1994

大学図書館所蔵 件 / 23



Includes bibliographical references and index



Providing the reader with an in-depth understanding of empirical inductive logic programming approaches - which can cope with imperfect data and can be used to construct knowledge bases for solving practical problems - this book also describes several practical applications in detail and gives an overview of other current applications of inductive logic programming. The book is at the leading edge of current research: inductive logic programming is an emerging field at the intersection of machine learning and logic programming. It proposes new methods for learning relational descriptions (dealing also with imperfect data) that can be viewed as alternative methods for logic program synthesis. The book also presents a veiw on inductive logic programming as a search of the structured space of logic program clauses, which addresses the issue of search complexity and search heuristics in detail. Two empirical inductive logic programming systems (LINUS and mFOIL) are described, as well as several applications of these systems.


  • Part 1 Empirical inductive logic programming: introduction
  • empirical ILP systems - an overview
  • LINUS - using attribute-value learners in an ILP framework
  • experiments in learning relations with LINUS
  • ILP as search for program clauses. Part 2 Learning relations from imperfect data: handling imperfect data in ILP
  • using heuristics to handle noise in ILP
  • mFOIL - extending noise-handling in FOIL
  • experiments in learning relations from noisy examples. Part 3 Applications of inductive logic programming: learning rules for early diagnosis of rheumatic diseases
  • finite element mesh design
  • an overview of selected ILP applications.

「Nielsen BookData」 より

関連文献: 1件中  1-1を表示


  • ISBN
    • 0134578708
  • LCCN
  • 出版国コード
  • タイトル言語コード
  • 本文言語コード
  • 出版地
    New York
  • ページ数/冊数
    xix, 293 p.
  • 大きさ
    25 cm
  • 分類
  • 件名
  • 親書誌ID