Ron Kohavi. Bottom-up induction of oblivious, read-once decision graphs. In Proceedings of the European Conference on Machine Learning, April 1994. Paper available by anonymous ftp from starry.stanford.edu:pub/ronnyk/euroML94.ps.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....ID3Inducer is a subclass of TDDTInducer that uses the ID3 algorithm to induce the decision tree [Qui86] OODGInducer (Oblivious, read Once Decision Graph inducer) is an abstract base class and a subclass of Inducer. OODGInducer uses a bottom up method for building levelled rooted decision graphs [Koh94]. If the environment variable OODGREMOVEDUP (which defaults to yes ) is set to yes, conflicting and duplicate instances are removed. num nodes( and num leaves( return the number of nodes and leaves respectively in the induced decision graph. HOODGInducer is a subclass of OODGInducer that ....

Ron Kohavi. Bottom-up induction of oblivious, read-once decision graphs. In Proceedings of the European Conference on Machine Learning, April 1994. Paper available by anonymous ftp from starry.stanford.edu:pub/ronnyk/euroML94.ps.

.... a decision tree as the start point, and transforms it into a RODG using the Minimum Description Length (MDL) principle [Rissanen, 1983] It is shown that the Smog algorithm achieves higher prediction accuracies than C4.5 in many artificial domains [Oliveira and Sangiovanni Vincentelli, 1995] Kohavi [1994] and Kohavi and Li [1995] explore Oblivious read Once Decision Graphs (OODG) that are similar to RODGs. All nodes at the same level of an OODG use the same test, so this type of decision graphs is oblivious. The term read Once , here, means that each attribute occurs at most once in each path of ....

....Oblivious read Once Decision Graphs (OODG) that are similar to RODGs. All nodes at the same level of an OODG use the same test, so this type of decision graphs is oblivious. The term read Once , here, means that each attribute occurs at most once in each path of an OODG. The Hoodg algorithm [Kohavi, 1994] constructs OODGs based on a set of ordered attributes using a bottom up method. The wrapper approach [John, Kohavi, and Pfleger, 1994] and or an entropy criterion are used to select an attribute subset from task supplied attributes and an ordering of selected attributes. Hoodg is shown to be ....

[Article contains additional citation context not shown here]

R. Kohavi, Bottom-up induction of oblivious, read-once decision graphs. Proceedings of the European Conference on Machine Learning, Berlin: Springer-Verlag, 154-169.

.... of top down approaches are top down induction of decision trees (TDIDT) algorithms (e.g. C4.5 (Quinlan, 1993) and unsupervised algorithms that use divisive methods to induce concept hierarchies (e.g. COBWEB (Fisher, 1989) Bottom up approaches include agglomerative algorithms (e.g. OODG (Kohavi, 1994)) Both types of approaches require information on the similarity among cases to generate hierarchies that accurately reflect the cases commonalities. For example, top down approaches require information on the similarity of the cases solutions, such as their topic and subtopic groupings. ....

Kohavi, R. (1994). Bottom-up induction of oblivious, read-once decision graphs. Proceedings of the European Conference on Machine Learning (pp. 154--169). Catania, Italy: Springer-Verlag.

No context found.

Kohavi, R. (1994a), Bottom-up induction of oblivious, read-once decision graphs, in "Proceedings of the European Conference on Machine Learning". Available by anonymous ftp from starry.Stanford.EDU:pub/ronnyk/euroML94.ps.

....we give comprehensibility an important weight. 3. It is compact. While related to comprehensibility, one does not imply the other. A perceptron (see below) might be a compact classifier, yet given an instance, it may be hard to understand the labelling process. Alternatively, a decision table (Kohavi 1995a) also see Chapter 5 on page 130) may be very large, yet labelling each instance is trivial: simply look it up in the table. Michie (1987) reported that when ID3 s output on the chess domain was shown to a domain expert, i.e. a chess master, it was completely opaque. Although it was very accurate, the tree was large, obscure, ....

Kohavi, R. (1994a), Bottom-up induction of oblivious, read-once decision graphs, in F. Bergadano & L. D. Raedt, eds, "Proceedings of the European Conference on Machine Learning", pp. 154--169.

....the performance never significantly degraded, an interesting phenomenon considering the fact that C4.5 is capable of locally discretizing features. 1 Introduction Many algorithms developed in the machine learning community focus on learning in nominal feature spaces (Michalski Stepp 1983, Kohavi 1994). However, many real world classification tasks exist that involve continuous features where such algorithms could not be applied unless the continuous features are first discretized. Continuous variable discretization has received significant attention in the machine learning community only ....

Kohavi, R. (1994), Bottom-up induction of oblivious, read-once decision graphs : strengths and limitations, in "Twelfth National Conference on Artificial Intelligence", pp. 613--618. Available by anonymous ftp from Starry.Stanford.EDU:pub/ronnyk/aaai94.ps.

No context found.

Kohavi, R. (1994a), Bottom-up induction of oblivious, read-once decision graphs, in "Proceedings of the European Conference on Machine Learning".

No context found.

Ron Kohavi. Bottom-up induction of oblivious, read-once decision graphs. In Proceedings of the European Conference on Machine Learning, April 1994. Paper available by anonymous ftp from starry.Stanford.EDU:pub/ronnyk/euroML94.ps.

....fragmentation problem causes partitioning of the data into fragments, when a high arity attribute is tested at a node. Both problems reduce the number of instances at lower nodes in the tree instances needed for statistical significance of tests performed during the tree construction process. In (Kohavi 1994), Oblivious read Once Decision Graphs (OODGs) were introduced as an alternative representation structure for supervised classification learning. OODGs retain most of the advantages of decision trees, while overcoming the two problems mentioned above. OODGs are similar to Ordered Binary Decision ....

.... OODGs are similar to Ordered Binary Decision Diagrams (OBDDs) Bryant 1986) which have been used in the engineering community to represent state graph models of systems, allowing verification of finite state systems with up to 10 120 states (Burch, Clarke, Long 1991) We refer the reader to (Kohavi 1994) for a discussion of related work. OODGs have a different bias from that of decision trees, and thus some concepts that are hard to represent as trees are easy to represent as OODGs, and vice versa. Since OODGs are graphs, they are easy for humans to perceive, and should be preferred over other ....

[Article contains additional citation context not shown here]

Kohavi, R. 1994. Bottom-up induction of oblivious, read-once decision graphs. In Proceedings of the European Conference on Machine Learning. Available by anonymous ftp from starry.Stanford.EDU:pub/ronnyk/euroML94.ps.

....of a discrete category label. Induction algorithms Induction algorithms induce categorizers. MLC currently provides a majority inducer, a nearest neighbor inducer [Aha92, AKA91] an ID3 like decision tree inducer [Qui86] HOODG for inducing oblivious read once decision graphs [Koh94b, Koh94a] Figure 1 illustrates the interactions between inducers, categorizers, testers, and wrappers. The decoupling of the inducer and the categorizer is critically important in supervised learning and plays an important role in the overall decomposition of the MLC library. 4.2 Overview of MLC ....

....which attribute to split on) at each node. ID3 inducer is a subclass of the top down decision tree inducer that uses the ID3 algorithm to induce the decision tree [Qui86] The OODG (Oblivious, read Once Decision Graph) inducer uses a bottom up method for building levelled rooted decision graphs [Koh94a, Koh94b] The HOODG inducer is a subclass of the OODG inducer that provides a hill climbing implementation of the OODG algorithm. A nearest neighbor inducer uses the IB algorithms described in [Aha92] Test result Test result is a class that provides information about the performance of a ....

[Article contains additional citation context not shown here]

Ron Kohavi. Bottom-up induction of oblivious, read-once decision graphs. In Proceedings of the European Conference on Machine Learning, April 1994. Paper available by anonymous ftp from starry.Stanford.EDU:pub/ronnyk/euroML94.ps.

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