Pneumonia identification using statistical feature selection

By Cosmin Adrian Bejan, Fei Xia, Lucy Vanderwende, M. Wurfel, Meliha Yetisgen-Yildiz. 2012

79 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Natural Language Processing for EHR-Based Computational Phenotyping (2018, background): “[97] Section headers, self-defined features, NegEx, and ConTex MetaMap Restricted set of time order physician daily note Uni-grams, bi-grams, UMLS concepts, assertion values associated with pneumonia expressions, statistical significance testing to rank features Carroll et al.”
Toward high-throughput phenotyping: unbiased automated feature extraction and selection from knowledge sources (2015, methodology): “used all possible unigrams, bigrams, and Unified Medical Language System (UMLS)(40) concepts to classify pneumonia with support vector machines (SVMs), with additional feature selection based on significance test scores obtained from the gold-standard labels.(41) Carroll et al.”
A systematic comparison of feature space effects on disease classifier performance for phenotype identification of five diseases (2015, methodology): “For example, Bejan et al. (2012) used a binary classifier for identifying pneumonia; comparing word n-grams, UMLS concepts and assertion values associated with pneumonia expressions.”