Directed Research Project Details
A semi-automatic framework for mining ERP patterns
| Author: | Jiawei Rong |
|---|---|
| Date: | November 28, 2006 |
| Time: | 16:30 |
| Location: | 220 Deschutes |
| Committee: | Dejing Dou (Chair) Allen Malony Don Tucker, Psychology |
Abstract
Event-related potentials (ERP) are created by averaging across segments of Electroencephalogram(EEG) data in different trials and time-locking to stimulus events or response. ERP data is a mixture of artifacts, noise, and components that are related to brain operations. In this paper, we propose a semi-automatic framework for mining ERP patterns. These patterns will be used as pattern elements for Neural ElectroMagnetic Ontology (NEMO) development. The framework includes temporal PCA decomposition of the raw ERP data, extraction of summary metrics, unsupervised learning of pattern clusters and supervised learning of classification rules. Some promising results are reported in this paper.