J. R. Quinlan. Discovering rules by induction from large collections of examples. In Expert Systems in the Micro-Electronic Age. Edinburgh University Press, 1979.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the overall knowledge of the retrieved and relevant image sets [6, 11, 12] This knowledge is captured in the form of a decision tree. A decision tree is a representation of a decision procedure for determining the classification of a given image. Quinlan s ID3 decision tree building algorithm [20, 21, 22] induces a decision tree for classifying the examples in the training set. Given the relevance judgment information, ID3 is used to build a decision tree based on the coherent part of CCV to classify images into relevant and non relevant sets. For our color based image retrieval, the class in ID3 ....

J. R. Quinlan. Discovering rules by induction from large collections of examples. Expert Systems in the Micro-electronic Age, pages 168-201, 1979.

....The main section of this article compares the model obtained direct induction of an EDAG for this problem with that obtained from human chess experts. Before this is done, the following section illustrates the induction of EDAGs for a simple chess problem. 2 Modeling a Simple Chess Dataset Quinlan (1979) describes ID3 models of 7 rook versus knight end game situations of increasing difficulty. The third problem involves 647 cases with 4 3 valued attributes, 3 2 valued attributes, and a 2 valued outcome. Figure 1 shows the decision tree induced by ID3 that solves this problem graphed as an EDAG in ....

Quinlan, J.R. (1979). Discovering rules by induction from large collections of examples. Michie, D., Ed. Expert Systems in the Micro Electronic Age. pp.168-201. Edinburgh, Edinburgh University Press.

....in C (Gaines, 1994) 3 Results with Some Standard Datasets In presenting a new algorithm for empirical induction it is appropriate first to illustrate its performance on test datasets published by others. Three examples are presented here: Cendrowska s (1987) contact lens example; one of Quinlan s (1979) chess end game examples also used by Cendrowska (1987) and Quinlan s (1987) prob disj example. The first two cases allow the ripple down rule solutions to be compared with conventional decision trees and production rules for simple deterministic data, and the last case allows the comparison to ....

.... = presbyopic prescription = myope lens = none prescription = myope tear production = normal lens = hard age = young tear production = normal lens = hard Figure 7 Normal and ripple down rules induced for contact lens dataset A somewhat more complex deterministic example is one of Quinlan s (1979) chess datasets that Cendrowska (1987) also analyzed with Prism. The data consists of 647 cases of a rook versus knight end game situation described in terms of four 3 valued and three 2 valued attributes leading to one of two conclusions. The ID3 tree for this data has 30 nodes. Prism s 15 rule ....

Quinlan, J.R. (1979). Discovering rules by induction from large collections of examples. Michie, D., Ed. Expert Systems in the Micro Electronic Age. pp.168-201. Edinburgh, Edinburgh University Press.

....be found from reference [3] This reference concentrates on the domain of pattern recognition, where features are usually numerical values. From the AI literature one can find methods to build decision trees from inputs with symbolic features. One such method is the ID3 (Iterative Dicothomizer 3) [17] based on the CLS (Concept Learning System) by [4] The algorithm starts from the empty root with all example cases. At each node one single feature is selected, and a new branch is generated corresponding to every different possible value of this feature. The examples remaining in the node are ....

J.R. Quinlan. Discovering rules by induction from large collections of examples. In Expert Systems in the Micro Electronic Age (ed. D.Michie). Edinburgh University Press, 1979.

....difficult objects to find) that subsequently discovered 100 of these rare stellar objects automatically previously only 10 had been known to exist. But how does a computer program discover the attributes influencing a given class outcome The original program to do this is called ID3 [23], its successor program C4.5 [24] Both were invented by an Australian Professor named Ross Quinlan, and nowadays are the most widely used machine learning program in knowledge and data discovery. For example in the robot data given in Table 1, the C4.5 program could be used to construct the ....

J R Quinlan. Discovering rules by induction from large collections of examples. In D Michie, editor, Expert Systems in the Microelectronic Age. Edinburgh University Press, 1979.

....and applied prediction systems to problems in such areas as agriculture[DM79] mathematics[Mit82] MUNB83] chemistry[BFL71] and industrial fuel production[Qui85] Unfortunately, the full range of applicability of these systems is unknown. Systematic evaluation. With few exceptions (e.g. [Qui79] [MC80] Qui85] CN86] QCHL86] Mic87] very little attention has been paid to the systematic evaluation of the systems using empirical data. Moreover, there has been very limited empirical validation of the effectiveness of proposed learning principles and approaches. Coupled with this is a ....

J. R. Quinlan. Discovering rules by induction from large collections of examples. In D. Michie, editor, Expert Systems in the MicroElectronic Age, pages 168--201, Edinburgh University Press, Edinburgh, 1979.

....Several lines of investigation of decision trees over finite information systems are known. These are test theory [5, 14, 16, 17, 31, 37, 38] theory of information systems and rough set theory [23, 24, 30] theory of questionnaires [25, 26] theory of decision tables [9] machine learning [27, 28, 29], search theory [1, 35] The terminology and methods of test theory, the groundwork for which was laid by [5, 37] will be used in the present paper for study of decision tree complexity as well as methods of rough set theory created in [23, 24, 30] 160 M.Moshkov Unimprovable Upper Bounds on ....

J.R. Quinlan, Discovering rules by induction from large collections of examples, in Experts Systems in the Microelectronic Age, Edited by D. Michie, Edinburg University Press (1979).

....literature and the scalability requirements of a data mining environment. The main insight, based on a careful analysis of the algorithms in the literature, is that most (to our knowledge, all) algorithms (including C4.5 [Qui93] CART [BFOS84] CHAID [Mag93] FACT [LV88] ID3 and extensions [Qui79, Qui83, Qui86, CFIQ88, Fay91] SLIQ and Sprint [MAR96, MRA95, SAM96] and QUEST [LS97] access the data using a common pattern, as described in Figure 1. We present data access algorithms that scale with the size of the database, adapt gracefully to the amount of main memory available, and are not ....

J.R. Quinlan. Discovering rules by induction from large collections of examples. In Expert Systems in the Micro Electronic Age, 1979.

....classification algorithms with data warehouse facilities are discussed in Section 2.3. 2.1 Scalable classification tree algorithms The machine learning community has proposed many algorithms for classification tree induction. Examples include Hunt s Concept Learning System [23] CART [3] ID3 [36, 37, 38, 12] and its extension to C4.5 [39] and FACT [28] Incremental versions of ID3 include ID4 [44] and ID5 [50] KATE [29] learns classification trees from complex structured data, while INFERULE [51] learns classification trees from inconclusive data. Classification tree algorithms that address ....

J. R. Quinlan. Discovering rules by induction from large collections of examples. In D. Michie, editor, Expert Systems in the Micro Electronic Age. Edinburgh University Press, 1979. BIBLIOGRAPHY 99

....data from previously measured or normative values. There are many different methodological approaches to data mining including machine learning, statistics, database oriented, etc. Machine learning approaches include learning from examples [64] conceptual clustering [66] decision tree induction [71], etc. Mathematical and statistical approaches include Bayesian inference [16, 17] and rough set [69] Database oriented approaches include attribute oriented induction [34, 32] Apriori [1, 4] etc. There are also other approaches including knowledge CHAPTER 1. INTRODUCTION 4 representation ....

J. R. Quinlan. Discovering rules by induction from large collections of examples. In D. Michie, editor, Expert Systems in the Micro Electronic Age. Edinburgh University Press, 1979.

....of induction is to use This work was carried out at the Speech Technology Research Laboratory, Department of Electrical Engineering, The University of Sydney. a known set of examples to a theory that explains both these examples and, hopefully, other unseen examples as well [6]. Since the most successful recognition systems are data driven, where the structure and characteristics of the speech signal is captured implicitly from the training data, this paper proposes a data driven knowledge based approach to consonant recognition in continuous speech. The system is based ....

Quinlan, J. R., "Discovering Rules by Induction from Large Collections of Examples", Machine Learning, Vol. 1, No. 1, 1979.

....Decision trees are typically generated from a set of examples of decisions. The essential characteristic of any such method is the attribute selection criterion used for choosing attributes to be assigned to the nodes of the decision tree being built. Such criteria include the entropy reduction [12, 13], the gini index of diversity [4] and others (e.g. 5, 6, 11) A decision tree decision structure can be an effective tool for describing a decision process, as long as all the required tests can be measured, and the decision making 2 situations it was designed for remain constant. Problems ....

Quinlan, J.R., "Discovering Rules By Induction from Large Collections of Examples", in D. Michie (Edr), Expert Systems in the Microelectronic Age, Edinburgh University Press, 1979.

....literature and the scalability requirements of a data mining environment. The main insight, based on a careful analysis of the algorithms in the literature, is that most (to our knowledge, all) algorithms (including C4.5 [Qui93] CART [BFOS84] CHAID [Mag93] FACT [LV88] ID3 and extensions [Qui79, Qui83, Qui86, CFIQ88, Fay91] SLIQ and Sprint [MAR96, MRA95, SAM96] and QUEST [LS97] access the data using a common pattern, as described in Figure 1. We present data access algorithms that scale with the size of the database, adapt gracefully to the amount of main memory available, and are not ....

J.R. Quinlan. Discovering rules by induction from large collections of examples. In Expert Systems in the Micro Electronic Age, 1979.

.... queries [16] and in sampling from labeled data [8, 25] Uncertainty sampling with a single classifier can also be viewed as a variation on the heuristic of training on misclassified instances [15, 33, 35] A familiar example of this is windowing, which appeared in Quinlan s first paper on ID3 [26], was questioned in [36] and re examined in Chapter 6 of the C4.5 book [27] As with uncertainty sampling, windowing builds a sequence of classifiers, selecting instances to add to the training set at each iteration. The key difference is its assumption that the class labels of all training ....

J. R. Quinlan. Discovering rules by induction from large collections of examples. In Expert systems in the micro-electronic age, Edinburgh, UK, 1979. Edinburgh University Press.

....be extracted from the decision tree. For example, the following classification rule can be extracted from the decision tree shown in Figure 2.3: IF size(x) medium AND transmission(x) automatic THEN mileage(x) medium. Quinlan uses entropy to induce decision trees in his ID3 algorithm [88]. Starting from an empty tree and a set of objects, ID3 chooses the attribute which generates maximum information gain (calculated from entropies) as the root node. A branch CHAPTER 2. RELATED WORK IN KDD 24 large weight auto manual medium low medium medium size light heavy medium transmission ....

J. R. Quinlan. Discovering rules by induction from large collections of examples. In D. Michie, editor, Expert Systems in the Micro Electronic Age. Edinburgh, England, 1979.

....tasks. and section 5 compares this performance with the performance of other reported recognisers. Finally, section 6 concludes this paper. 2. INDUCTIVE SYSTEMS Inductive systems have already been used to extract classification knowledge from large databases and collections of examples [15, 16, 17] The essence of induction is to use a known set of examples and match it to a theory that explains both these examples and, hopefully, other unseen examples as well [18] The inductive tool used in this paper is an implementation of C4.5, a descendant of IDZ, which in turn is based on ....

....training data, about 36 items are misclassified at that leaf. This indicates a classification error rate of 11.6 at that leaf. This process continues until all the leaves are classified. Figure 2 shows a selection of pruned rules generated from the unpruned decision tree shown in Figure 1. Rule 151: Rule 217: trajF1 = N trajF2 = L trajF5 = R freqF1 = 0 freqF2 = 0 freqF2 1636 class N [77.8 ] freqF4 0 class m [83.3 ] Rule 79: Rule 185: trajF1 = L trajF4 = N trajF2 = F trajF5 = L trajF3 = N t class n [82.4 ] freqF1 496.4 freqF2 = 1169.7 Rule 175: tfreqF5 = 0 trajF5 = F ....

[Article contains additional citation context not shown here]

J. R. Quinlan (1979). Discovering rules by induction from large collections of examples, in Expert Systems in Micro Electronic Age, D. Mirchie, ed. Edinburgh: Edinburgh University Press.

....this century. 3 Hunt was one of the early pioneers who modelled a theory of human concept learning using computer programs. He developed a series of algorithms called concept learning systems (CLS 1 to CLS 9) described in Hunt (1962) and Hunt et al. 1966) Quinlan s well known ID3 algorithm (Quinlan, 1979), was descended from these systems. Basically, ID3 was a procedure for discriminating between two classes in domains which were entirely free from uncertainty. In fact, ID3 was developed initially for performing chess endgame analysis, discriminating between winning and non winning positions) ....

Quinlan, J.R. (1979). Discovering rules by induction from large collections of examples. In Expert Systems in the Micro-Electronic Age, edited by D. Michie, pp. 168--201. Edinburgh: Edinburgh University Press.

....are often directly incorporated into many machine learning algorithms. Statistical analysis is commonly the basis for optimising decision tree rule construction algorithms; for example, Quinlan s ID3 uses chi square tests to decide whether an attribute should be added to a classification tree (Quinlan, 1979). Once a model has been produced, it may be over fitted to the training data (ie, it contains overspecific rules that produce accurate classification results on the training data, but does not contain sufficient generalisations to perform adequately on new data) To overcome this problem, methods ....

Quinlan, J.R. (1979) "Discovering rules by induction from large collections of examples", in Expert Systems in the Micro-Electronic Age, edited by D.

....with four other algorithms in three medical domains. Firstly, we give a brief description of the algorithms used for comparison. Secondly, details of the medical domain are given and evaluation criteria presented. 3.1. Comparative Algorithms 3.1.1. Assistant Assistant [13] is a descendant of ID3 [24] and CLS [25] Assistant induces rules in the form of decision trees. The entropy measure is used to guide the growth of the decision tree, as described in [1] In addition, Assistant can apply a tree pruning method based on a technique of maximal classification precision. This technique detects ....

Quinlan J. (1979) Discovering rules by induction from large collections of examples Introductory readings in expert systems Ed. D. Michie, London: Gordon and Breach pp.33-46.

No context found.

J. R. Quinlan. Discovering rules by induction from large collections of examples. In Expert Systems in the Micro-Electronic Age. Edinburgh University Press, 1979.

No context found.

Quinlan J. R. Discovering rules by induction from large collections of examples. In D. Michie (Ed.), Expert systems in the micro electronic age. Edinburgh Univ. Press, 1979.

No context found.

J. R. Quinlan. Discovering rules by induction from large collections of examples. In Expert Systems in the Micro-Electronic Age. Edinburgh University Press, 1979.

No context found.

Quinlan J. R. Discovering rules by induction from large collections of examples. In D. Michie (Ed.), Expert systems in the micro electronic age. Edinburgh Univ. Press, 1979.

No context found.

Quinlan J. R., Discovering Rules by Induction from Large Collection of Examples, Knowledge-base systems in the Micro Electronic Age, Edinburgh University Press, 1979.

No context found.

Quinlan, J.R. : "Discovering rules by induction from large collections of examples", in Expert Systems in the Micro-electronic Age, Michie,D. (ed.), Edinburgh University Press, 1979.

First 50 documents

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC