We present algorithms for the generation of uniformly distributed Bayesian networks with constraints on induced width. The algorithms use ergodic Markov chains to generate samples....
Methods for imputation of missing data in the so-called least-squares approximation approach, a non-parametric computationally efficient multidimensional technique, are experiment...
—In this paper, we propose a novel method for extracting handwritten characters from multi-language document images, which may contain various types of characters, e.g. Chinese, ...
Yonghong Song, Guilin Xiao, Yuanlin Zhang, Lei Yan...
Many emerging large-scale data science applications require searching large graphs distributed across multiple memories and processors. This paper presents a distributed breadthï¬...
Andy Yoo, Edmond Chow, Keith W. Henderson, Will Mc...
The problem of estimating the kth frequency moment Fk for any nonnegative k, over a data stream by looking at the items exactly once as they arrive, was considered in a seminal pap...