background and objectives: accurate detection of type and severity of hepatitis is crucial for effective treatment of the disease. while several computational algorithms for detection of hepatitis have been proposed to date, their limited performance leaves room for further improvement. this paper proposes a novel computational method for the diagnosis of hepatitis b using pattern detection tec...

Multiple classifier system exhibits strong classification capacity compared with single classifiers, but they require significant computational resources. Selective ensemble aims to attain equivalent or better accuracy fewer classifiers. However, current methods fail identify precise solutions for constructing an classifier. In this study, we propose design technique based on the perturbation b...

We present a simple semi-supervised learning algorithm for named entity recognition (NER) using conditional random fields (CRFs). The algorithm is based on exploiting evidence that is independent from the features used for a classifier, which provides high-precision labels to unlabeled data. Such independent evidence is used to automatically extract highaccuracy and non-redundant data, leading ...