“…The traffic dataset [21], [22] describes the task of detecting sections of roads where a traffic problem-an accident or a congestion-has occurred at a specific time.…”
Section: Relational Datasetsmentioning
confidence: 99%
“…In [21] and [22], a discretization provided by experts in the field was used for the three numerical arguments of the traffic dataset. Using the same discretization, ECL-GSD obtained results that are slightly superior to those obtained using Fayyad and Irani algorithm [on the accidents dataset, the average accuracy on the test and training sets is 0.92 (0.03) and 0.94 (0.02) and the average simplicity is 5.10 (0.93).…”
This paper analyzes experimentally discretization algorithms for handling continuous attributes in evolutionary learning. We consider a learning system that induces a set of rules in a fragment of first-order logic (evolutionary inductive logic programming), and introduce a method where a given discretization algorithm is used to generate initial inequalities, which describe subranges of attributes' values. Mutation operators exploiting information on the class label of the examples (supervised discretization) are used during the learning process for refining inequalities. The evolutionary learning system is used as a platform for testing experimentally four algorithms: two variants of the proposed method, a popular supervised discretization algorithm applied prior to induction, and a discretization method which does not use information on the class labels of the examples (unsupervised discretization). Results of experiments conducted on artificial and real life datasets suggest that the proposed method provides an effective and robust technique for handling continuous attributes by means of inequalities.
“…The traffic dataset [21], [22] describes the task of detecting sections of roads where a traffic problem-an accident or a congestion-has occurred at a specific time.…”
Section: Relational Datasetsmentioning
confidence: 99%
“…In [21] and [22], a discretization provided by experts in the field was used for the three numerical arguments of the traffic dataset. Using the same discretization, ECL-GSD obtained results that are slightly superior to those obtained using Fayyad and Irani algorithm [on the accidents dataset, the average accuracy on the test and training sets is 0.92 (0.03) and 0.94 (0.02) and the average simplicity is 5.10 (0.93).…”
This paper analyzes experimentally discretization algorithms for handling continuous attributes in evolutionary learning. We consider a learning system that induces a set of rules in a fragment of first-order logic (evolutionary inductive logic programming), and introduce a method where a given discretization algorithm is used to generate initial inequalities, which describe subranges of attributes' values. Mutation operators exploiting information on the class label of the examples (supervised discretization) are used during the learning process for refining inequalities. The evolutionary learning system is used as a platform for testing experimentally four algorithms: two variants of the proposed method, a popular supervised discretization algorithm applied prior to induction, and a discretization method which does not use information on the class labels of the examples (unsupervised discretization). Results of experiments conducted on artificial and real life datasets suggest that the proposed method provides an effective and robust technique for handling continuous attributes by means of inequalities.
“…, * varn ∈ i (ie. Zj = {k ∈ ⊥ | k consumes * varj ∈ i}) then: [2,15], [3,14], [3,15], [1,13], [1,16], [4,13], [4,16]}. Suppose that we are looking for a two-literal clause.…”
Section: Examplementioning
confidence: 99%
“…The main advantage of using the macro-based refinement operator is the reduction of the search space; however, there is a cost for obtaining the macro's set D. To analyze the performance of the macro-based method, we perform experiments on four datasets. The first dataset contains 180 positive and 17 negative examples of valid chess moves for five pieces 8 ; the second consists of 3340 positive and 1498 negative examples of "safe" 9 minesweeper moves; the third one is the dataset used in [2] with 256 positive and 512 negative examples of road sections where a traffic problem has occurred; and the last one is the ILP benchmark dataset mutagenesis [12] with 125 positive and 63 negative examples.…”
Abstract. Refinement operators are frequently used in the area of multirelational learning (Inductive Logic Programming, ILP) in order to search systematically through a generality order on clauses for a correct theory. Only the clauses reachable by a finite number of applications of a refinement operator are considered by a learning system using this refinement operator; ie. the refinement operator determines the search space of the system. For efficiency reasons, we would like a refinement operator to compute the smallest set of clauses necessary to find a correct theory. In this paper we present a formal method based on macro-operators to reduce the search space defined by a downward refinement operator (ρ) while finding the same theory as the original operator. Basically we define a refinement operator which adds to a clause not only single-literals but also automatically created sequences of literals (macro-operators). This in turn allows us to discard clauses which do not belong to a correct theory. Experimental results show that this technique significantly reduces the search-space and thus accelerates the learning process.
We describe a methodology for upgrading existing attribute value learners towards rst order logic. This method has several advantages: one can pro t from existing research on propositional learners (and inherit its e ciency and e ectiveness), relational learners (and inherit its expressiveness) and PAC-learning (and inherit its theoretical basis). Moreover there is a clear relationship between the new relational system and its propositional counterpart. This makes the ILP system easy to use and understand by users familiar with the propositional counterpart. We demonstrate the methodology on the ICL system which is an upgrade of the propositional learner CN2.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.