10 The Author 13 Acknowledgements 15 1 Introduction 16 1.1 Definition of Learning : : : : : : : : : : : : : : : : : : : : : : : : 16 1.2 The objectives of ML : : : : : : : : : : : : : : : : : : : : : : : : : 17 1.3 Approaches taken so far : : : : : : : : : : : : : : : : : : : : : : : 18 1.4 Motivation for the project : : : : : : : : : : : : : : : : : : : : : : 20 1.5 The Structure of the Thesis : : : : : : : : : : : : : : : : : : : : : 21 2 Theory of Inductive Learning 22 2.1 Introduction : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 22 2.2 Induction as a Search : : : : : : : : : : : : : : : : : : : : : : : : : 23 2.2.1 The Goal: Hypothesis : : : : : : : : : : : : : : : : : : : : 24 2.2.2 The Search Space: Hypothesis Space : : : : : : : : : : : : 24 2.2.3 The operators : : : : : : : : : : : : : : : : : : : : : : : : : 26 2.3 Approaches : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 27 2.3.1 Statistical Classification : : : : : : : : : : : : : : : : : : : 27...

This report describes experiments in learning models for basic flight manoeuvres from behavioural traces of a human pilot when using a flight simulator. A first set of experiments using decision trees is presented. The auto-pilot built with the generated decision trees flies more smoothly than the human pilot. However the results show also that propositional logic-level representations, like decision trees, are inadequate to fully solve the problem. A learning system using a first-order representation is required. However, current Inductive Logic Programming systems have severe limitations when dealing with such complex domains due to inefficiencies of searching large hypothesis spaces. An important issue to make the hypothesis space search tractable and efficient is the use of background knowledge. Some first results are reported based on a system under development that already shows some uses of background knowledge at a "local" level of learning a single predicate. Identification of...

Abstract not found

The use of expert systems is becoming more and more widespread, making the need for appropriate machine learning techniques more acute to help ease the knowledge aquisition bottleneck. Additionally, the increasing number of large databases offers a vast potential for the automatic generation of new knowledge by machines and its communication to people in a comprehensible form. In response to these events, this paper provides an overview of current machine learning work with a particular emphasis on rule induction techniques. Firstly we provide a summary of existing rule induction techniques, including descriptions of the ID3 and AQ algorithms. Secondly, we review recent developments in rule induction technology which overcome some of the practical limitations of these basic algorithms including noise handling, probabilistic classification, large data sets and incremental learning. Finally, we describe the state of current research in machine learning and the directions in which it is heading, addressing the difficult problems of constructive induction and representation change.

. When attempting to discover by learning concepts embedded in data, it is not uncommon to find that information is missing from the data. Such missing information can diminish the confidence on the concepts learned from the data. This paper describes a new approach to fill missing values in examples provided to a learning algorithm. A decision tree is constructed to determine the missing values of each attribute by using the information contained in other attributes. Also, an ordering for the construction of the decision trees for the attributes is formulated. Experimental results on three datasets show that completing the data by using decision trees leads to final concepts with less error under different rates of random missing values. The approach should be suitable for domains with strong relations among the attributes, and for which improving accuracy is desirable even if computational cost increases. 1 Introduction Machine learning techniques have been successfully...

www.fmt.vein.hu/softcomp Abstract. Fuzzy decision tree induction algorithms require the fuzzy quantization of the input variables. This paper demonstrates that supervised fuzzy clustering combined with similarity-based rule-simplification algorithms is an effective tool to obtain the fuzzy quantization of the input variables, so the synergistic combination of supervised fuzzy clustering and fuzzy decision tree induction can be effectively used to build compact and accurate fuzzy classifiers.

The decision tree induction is a widely applied technique in machine learning. Algorithms based on such technique with careful implementation can reach the computational complexity # n log n). However, a special treatment of decision trees is needed in case of missing values.

Abstract. In this paper we present a new approach to handling incomplete information and classifier complexity reduction. We describe a method, called D 3 RJ, that performs data decomposition and decision rule joining to avoid the necessity of reasoning with missing attribute values. In the consequence more complex reasoning process is needed than in the case of known algorithms for induction of decision rules. The original incomplete data table is decomposed into sub-tables without missing values. Next, methods for induction of decision rules are applied to these sets. Finally, an algorithm for decision rule joining is used to obtain the final rule set from partial rule sets. Using D 3 RJ method it is possible to obtain smaller set of rules and next better classification accuracy than standard decision rule induction methods. We provide an empirical evaluation of the D 3 RJ method accuracy and model size on data with missing values of natural origin. 1

Keywords: In this chapter methods of handling missing attribute values in data mining are described. These methods are categorized into sequential and parallel. In sequential methods, missing attribute values are replaced by known values first, as a preprocessing, then the knowledge is acquired for a data set with all known attribute values. In parallel methods, there is no preprocessing, i.e., knowledge is acquired directly from the original data sets. In this chapter the main emphasis is put on rule induction. Methods of handling attribute values for decision tree generation are only briefly summarized. Missing attribute values, lost values, do not care conditions, incomplete data, imputation, decision tables. 1.

Oscar Ortega Lobo oortega@udea.edu.co Dept. of Systems Engineering, Faculty of Engineering, University of Antioquia, Medellin P.O. BOX 1226, Colombia Masayuki Numao numao@cs.titech.ac.jp Dept. of Computer Science, Graduate School of Information Sciences and Engineering, Tokyo Institute of Technology, Tokyo 152-8552, Japan Abstract In empirical research on machine learning, usually an algorithm is chosen among a set of algorithms, by using as a selection criteria the average of an indicator of algorithm performance. Carefully controlled experiments are conducted on a nite number of domains for which the same learning task is applicable, and the average of the performance indicator is calculated. The best-on-average selection criteria ignores the fact that for some of the learning tasks the performance of the selected algorithm is not the best among the candidate algorithms. Instead of the previous kind of selection criteria, it could be more informative, for both the...