Context-dependent deep learning
July 1, 2021
Explicitly representing an agent’s context has been shown to have many benefits, which should also apply to machine learning. In this paper, we describe an approach to do this called context-dependent deep learning (CDDL), which is based on earlier work in context-mediated behavior (CMB) that uses contextual schemas (c-schemas) to represent clas-ses of situations along with knowledge useful in them. These are then recalled, and they guide reasoning in the corre-sponding contexts. CDDL stores knowledge about deep neural network structure and weights in c-schemas, which al-lows context-specific learning. Our work is being developed in the domain of seabird detection in aerial images of islands for use by biologists.
Citation Information
Publication Year | 2021 |
---|---|
Title | Context-dependent deep learning |
DOI | 10.21494/ISTE.OP.2021.0690 |
Authors | Roy M. Turner, Cyndy Loftin, Alex Revello, Logan R. Kline, Meredith Lewis, Salimeh Yasai–Sekeh |
Publication Type | Article |
Publication Subtype | Journal Article |
Series Title | Modeling and Using Context |
Index ID | 70229733 |
Record Source | USGS Publications Warehouse |
USGS Organization | Coop Res Unit Leetown |