Data Mining with Bayesian Network learning has two important characteristics: under broad conditions learned edges between variables correspond to causal influences, and second, f...
Ioannis Tsamardinos, Constantin F. Aliferis, Alexa...
Statistical language modeling has been successfully used for speech recognition, part-of-speech tagging, and syntactic parsing. Recently, it has also been applied to information r...
A method for color indexing is proposed that is based upon nonparametric statistical techniques. Nonparametrics compare the ordinal rankings of sample populations, and maintain the...
The ratio of two probability densities can be used for solving various machine learning tasks such as covariate shift adaptation (importance sampling), outlier detection (likeliho...
Abstract. The Gram matrix plays a central role in many kernel methods. Knowledge about the distribution of eigenvalues of the Gram matrix is useful for developing appropriate model...