A parallel implementation of the specialized interior-point algorithm for multicommodity network flows introduced in [5] is presented. In this algorithm, the positive definite syst...
The options framework provides a method for reinforcement learning agents to build new high-level skills. However, since options are usually learned in the same state space as the...
Graph-based semi-supervised learning has recently emerged as a promising approach to data-sparse learning problems in natural language processing. All graph-based algorithms rely ...
The problem of identifying mislabeled training examples has been examined in several studies, with a variety of approaches developed for editing the training data to obtain better...
We revise the classic methodology to find the multi-product break-even point. In the current paper we propose a solution to the problem under uncertainty conditions, based on DurÃ...