The probability that a term appears in relevant documents ( ) is a fundamental quantity in several probabilistic retrieval models, however it is difficult to estimate without rele...
We develop a method for predicting query performance by computing the relative entropy between a query language model and the corresponding collection language model. The resultin...
—This paper combines artificial neural networks (ANN), fuzzy optimization and time-series econometric models in one unified framework to form a hybrid intelligent early warning...
Recently, due to problems arising from lattice mismatch in thin film growth in semiconductor manufacturing industry, researchers have shown great interest in modeling the physioch...
The main statistics used in rough set data analysis, the approximation quality, is of limited value when there is a choice of competing models for predicting a decision variable. ...