905 search results - page 152 / 181 » Active Learning for Pipeline Models |
CSREAESA
2004
13 years 9 months ago
2004
- The model of a simple perceptron using phase-encoded inputs and complex-valued weights is presented. Multilayer two-input and three-input complex-valued neurons (CVNs) are implem...
ESANN
2001
13 years 9 months ago
2001
Abstract. Sigmoidal or radial transfer functions do not guarantee the best generalization nor fast learning of neural networks. Families of parameterized transfer functions provide...
NAACL
2004
13 years 9 months ago
2004
We present a technique for augmenting annotated training data with hierarchical word clusters that are automatically derived from a large unannotated corpus. Cluster membership is...
WOA
2007
13 years 9 months ago
2007
Abstract— Situated agents engaged in open systems continually face with external events requiring adequate services and behavioral responses. In these conditions agents should be...
ANOR
2010
13 years 8 months ago
2010
Abstract. We consider a continuous-time model for inventory management with Markov modulated non-stationary demands. We introduce active learning by assuming that the state of the ...