172 search results - page 5 / 35 » Learning predictive representations from a history |
KDD
2005
ACM
14 years 27 days ago
2005
ACM
Mining data streams is important in both science and commerce. Two major challenges are (1) the data may grow without limit so that it is difficult to retain a long history; and (...
TSE
2008
13 years 7 months ago
2008
This paper introduces a new technique for predicting latent software bugs, called change classification. Change classification uses a machine learning classifier to determine wheth...
COLING
2010
13 years 2 months ago
2010
The problem addressed in this paper is to predict a user's numeric rating in a product review from the text of the review. Unigram and n-gram representations of text are comm...
ICAC
2006
IEEE
14 years 1 months ago
2006
IEEE
Abstract— Shared computing utilities allocate compute, network, and storage resources to competing applications on demand. An awareness of the demands and behaviors of the hosted...
BMVC
2001
13 years 9 months ago
2001
Visual interpretation of events requires both an appropriate representation of change occurring in the scene and the application of semantics for differentiating between different...