Lazy Learning by David W. Aha (auth.), David W. Aha (eds.)

By David W. Aha (auth.), David W. Aha (eds.)

This edited assortment describes contemporary growth on lazy studying, a department of desktop studying relating algorithms that defer the processing in their inputs, respond to details requests via combining saved facts, and usually discard developed replies. it's the first edited quantity in AI in this subject, whose many synonyms contain `instance-based', `memory-based'. `exemplar-based', and `local learning', and whose subject intersects case-based reasoning and edited k-nearest neighbor classifiers. it truly is meant for AI researchers and scholars drawn to pursuing fresh growth during this department of desktop studying, yet, as a result breadth of its contributions, it's going to additionally curiosity researchers and practitioners of information mining, case-based reasoning, statistics, and development recognition.

Show description

Read or Download Lazy Learning PDF

Best nonfiction_8 books

Wave Phenomena: Theoretical, Computational, and Practical Aspects

IJ:1 June of 1987 the guts for utilized arithmetic and desktop technological know-how at San Jose country college obtained a bequest of over part one million money from the property of Mrs. Marie Woodward. within the commencing article of this choice of papers Jane Day, the founding father of the heart, describes the history that ended in this present.

Ion Flux in Pulmonary Vascular Control

6 Ions can go through a unmarried membrane channel at a expense of 10 ions/second. during the last decade the facility to degree ion flux so accurately and to record the hole and shutting of person ion channels has supplied a strong device to these engaged on delicate muscle body structure and vascular reactivity.

Selforganization: Portrait of a Scientific Revolution

Can be advanced with out with the ability to get replaced by means of whatever »still extra simple«. This turned obtrusive with the aid of computing device versions of deterministic-recursive platforms within which basic mathematical equation platforms supply a really advanced habit. (2) Irregularity of nature isn't really taken care of as an anomaly yet turns into the focal point of study and therefore is said to be basic.

Extra resources for Lazy Learning

Example text

Weights that go to infinity when a query equals a stored data point allow exact interpolation of the stored data. Finite weights lead to smoothing of the data. Weight functions that go to zero at a finite distance allow faster implementations, since points further from the query than that distance can be ignored with no error. As mentioned previously, kernels with a fixed finite radius raise the possibility of not having enough or any points within the non-zero area, a possibility that must be handled by the locally weighted learning system.

46) This expression of the prediction intervals is independent of the output values of the training data Yi, and reflects how well the data is distributed in the input space. However, the variance only reflects the difference between the prediction and the mean prediction, and not the difference between the prediction and the true value, which requires knowledge of the predictor's bias. Only when the local model structure is correct will the bias be zero. To conveniently derive an estimate of 0-2(X) we will define some additional quantities in terms of the weighted variables.

In that case we argue that one should take advantage of the opportunity to locally weight the training procedure, and get better performance (Vapnik 1992; Bottou and Vapnik 1992; Vapnik and Bouou 1993). 50 CHRISTOPHER G. ATKESON 14. Implementing Locally Weighted Learning There are several concerns about locally weighted learning systems, including whether locally weighted learning systems can answer queries fast enough and whether their speed will unacceptably degrade as the size of the database grows.

Download PDF sample

Rated 4.82 of 5 – based on 49 votes