Extensions to the CART algorithm

Crawford, Stuart L.

doi:10.1016/0020-7373(89)90027-8

Cited by 121 publications

(47 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Just as many training parameters in an artificial neural network must be determined empirically for a particular application [54], these parameters for a probabilistic network can be determined empirically using some training and testing data. A decision tree can be used as a classifier or to reveal the underlying patterns of data [59][60][61]. As a classifier, a decision tree is used to divide a problem region into several sub-regions with all the examples (data points) in one sub-region having the same target value.…”

Section: -4 Results and Discussionmentioning

confidence: 99%

See 1 more Smart Citation

The Monitoring, Detection, Isolation and Assessment of Information Warfare Attacks Through Multi-Level, Multi-Scale System Modeling and Model Based Technology

Ye¹

2004

View full text Add to dashboard Cite

REPORT DOCUMENTATION PAGE Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden to

show abstract

Section: -4 Results and Discussionmentioning

confidence: 99%

“…CHAID uses a chi-square test to find the most significant split points on the predictor variables. GINI is essentially CART [59][60][61] with the GINI index. The GINI impurity index is defined as the possibility at a test node that a randomly chosen case is classified incorrectly.…”

Section: -4 Results and Discussionmentioning

confidence: 99%

The Monitoring, Detection, Isolation and Assessment of Information Warfare Attacks Through Multi-Level, Multi-Scale System Modeling and Model Based Technology

Ye¹

2004

View full text Add to dashboard Cite

show abstract

“…An early incremental decision tree algorithm was proposed by Crawford (1989) based on CART (Breiman et al 1984). When a new training instance would cause a new test to be picked at a decision node, the entire subtree rooted at this node is discarded and rebuilt based on the corresponding subset of the training examples.…”

Section: Incremental Decision Treesmentioning

confidence: 99%

Interactively Training Pixel Classifiers

Piater

Riseman

Utgoff

1999

Int. J. Patt. Recogn. Artif. Intell.

View full text Add to dashboard Cite

Manual generation of training examples for supervised learning is an expensive process. One way to reduce this cost is to produce training instances that are highly informative. To this end, it would be beneficial to produce training instances interactively. Rather than provide a supervised learning algorithm with one complete set of training examples before learning commences, it would be better to produce each new training instance based on knowledge of which instances the learner would otherwise misclassify. Whenever the learner receives one or more new training examples, it should update its classifier incrementally and, in real time, provide the teacher with feedback about its current performance. The feasibility of such an approach is demonstrated on a realistic image pixel classification task. Here, the number of training instances involved in building a classifier was reduced by several orders of magnitude, at no perceivable loss of classification accuracy.

show abstract

“…Schlimmer and Fisher's (1986) ID4 demonstrated incremental tree induction through test revision and discarding of subtrees. Crawford (1989) has constructed an incremental version of the CART algorithm (Breiman, Friedman, Olshen & Stone, 1984). When a new example is received, if a new test would be picked at a decision node, a new subtree with the new test is constructed by building the new subtree from scratch from the corresponding subset of the training examples.…”

Section: Related Workmentioning

confidence: 99%

Untitled

1997

View full text Add to dashboard Cite

Abstract. The ability to restructure a decision tree efficiently enables a variety of approaches to decision tree induction that would otherwise be prohibitively expensive. Two such approaches are described here, one being incremental tree induction (ITI), and the other being non-incremental tree induction using a measure of tree quality instead of test quality (DMTI). These approaches and several variants offer new computational and classifier characteristics that lend themselves to particular applications.

show abstract

Extensions to the CART algorithm

Cited by 121 publications

References 4 publications

The Monitoring, Detection, Isolation and Assessment of Information Warfare Attacks Through Multi-Level, Multi-Scale System Modeling and Model Based Technology

The Monitoring, Detection, Isolation and Assessment of Information Warfare Attacks Through Multi-Level, Multi-Scale System Modeling and Model Based Technology

Interactively Training Pixel Classifiers

Untitled

Contact Info

Product

Resources

About