We present a unified view of two state-of-theart non-projective dependency parsers, both approximate: the loopy belief propagation parser of Smith and Eisner (2008) and the relaxe...
The Viterbi algorithm is an efficient and optimal method for decoding linear-chain Markov Models. However, the entire input sequence must be observed before the labels for any tim...
It is often expensive to acquire data in real-world data mining applications. Most previous data mining and machine learning research, however, assumes that a fixed set of trainin...
One purpose of empirical software engineering is to enable an understanding of factors that influence software development. Surveys are an appropriate empirical strategy to gather...
Teade Punter, Marcus Ciolkowski, Bernd G. Freimut,...
— In this paper, we present a new approach for automatic car detection from aerial images. The system exploits a robust machine learning method known as boosting for efficient c...
Thuy Thi Nguyen, Helmut Grabner, Horst Bischof, B....