Splitting and Merging Version Spaces to Learn Disjunctive Concepts
نویسندگان
چکیده
ÐWe have modified the original version space strategy in order to learn disjunctive concepts incrementally and without saving past training instances. The algorithm time complexity is also analyzed, and its correctness is proven. Index TermsÐDisjunctive concept, incremental learning, merge, split, version space.
منابع مشابه
The Iterated Version Space Learning
Inspired with Version Space learning the Iterated Version Space Algorithm IVSA has been designed and implemented to learn disjunctive concepts IVSA dynamically partitions its search space of potential hypotheses of the target concept into contour shaped regions until all training instances are maximally correctly classi ed
متن کاملA Symmetric Version Space Algorithm for Learning Disjunctive String Concepts
The symmetric version space algorithm (SVS) learns disjunctions of string patterns by example. The learnable string concepts are a subset of regular expressions. The running time of the algorithm is reduced, because the system learns a top{down description of the string concepts. Di erent parts of the algorithm learn descriptions at di erent levels of the concept independently. This technique i...
متن کاملThe Iterated Version Space Algorithm
We present the Iterated Version Space Al gorithm IVSA which retains many advan tages of the Version Space Algorithm while handling disjunctive concepts and noise IVSA repeatedly generates hypotheses for regions of the concept space and combines these regional hypotheses to produce an overall concept hypothesis Experiments were conducted to learn English pronunci ation rules from word pronunciat...
متن کاملThe Common Order-Theoretic Structure of Version Spaces and ATMS's
This paper exposes the common order-theoretic properties of the structures manipulated by the version space algorithm [Mit78]and the assumption-based truth maintenance systems (ATMS) [dk86a,dk86b] by recasting them in the framework of convex spaces. Our analysis of version spaces in this framework reveals necessary and sufficient conditions for ensuring the preservation of an essential finite r...
متن کاملPU-learning disjunctive concepts in ILP
PU learning is a variant of semi-supervised learning where labels of only one class are given. In his paper, we present an ILP method that PU-learns disjunctive concepts. The problem is motivated by some work on natural language learning, more specifically, learning the meaning of n-grams from (sentence, context)-pairs, where the context is described using first-order logic. An earlier solution...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- IEEE Trans. Knowl. Data Eng.
دوره 11 شماره
صفحات -
تاریخ انتشار 1999