Abstract
In this paper, we develop a new neural model that deals with continuation value of inputs for some practical applications of pattern recognition task. An essential core of the model is use of a novel vector representation of a target concept in a multi-level informational hierarchy that makes the model possess category formation ability from incomplete observation of the target. Simulation results demonstrate the usefulness of the model for a facial image recognition task, even if it is carried out under an incremental and unsupervised learning environment.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the SICE Annual Conference |
| Pages | 1483-1487 |
| Number of pages | 5 |
| Publication status | Published - 2004 |
| Event | SICE Annual Conference 2004 - Sapporo, Japan Duration: 2004 Aug 4 → 2004 Aug 6 |
Other
| Other | SICE Annual Conference 2004 |
|---|---|
| Country/Territory | Japan |
| City | Sapporo |
| Period | 04/8/4 → 04/8/6 |
Keywords
- Concept formation
- Hebbian rule
- Incremental learning
- Neural networks
- Pattern recognition
- Self-organization
ASJC Scopus subject areas
- Engineering(all)