Publication: Feature Selection for Privileged Modalities in Disease Classification
Loading...
Date
Advisor
Coadvisor
Graduate program
Undergraduate course
Journal Title
Journal ISSN
Volume Title
Publisher
Type
Work presented at event
Access right
Abstract
Multimodal data allows supervised learning while considering multiple complementary views of a problem, improving final diagnostic performance of trained models. Data modalities that are missing or difficult to obtain in clinical situations can still be incorporated into model training using the learning using privileged information (LUPI) framework. However, noisy or redundant features in the privileged modality space can limit the amount of knowledge transferred to the diagnostic model during the LUPI learning process. We consider the problem of selecting desirable features from both standard features which are available during both model training and testing, and privileged features which are only available during model training. A novel filter feature selection method named NMIFS+ is introduced that considers redundancy between standard and privileged feature spaces. The algorithm is evaluated on two disease classification datasets with privileged modalities. Results demonstrate an improvement in diagnostic performance over comparable filter selection algorithms.
Description
Keywords
Clinical decision support, Feature selection, Knowledge transfer, Multimodal data, Mutual information, Privileged learning
Language
English
Citation
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), v. 13050 LNCS, p. 69-80.
