Name
Title | ||
|---|---|---|
Examining Hierarchy and Granularity of Classifiers in Compatibility-based Classifier Personalization. |
Abstract | ||
|---|---|---|
We have been investigating a method of personalizing activity recognition system based on compatibility with pooled classifiers. In this paper, we examine two factors that may affect the effectiveness of the compatibility-based method over traditional “one classifier fits all (OFA)” approach: The hierarchy of classifiers and the granularity of classifiers in compatibility calculation. We confirmed that formations of activity group in the hierarchical model exist that drastically improve the performance over OFA approach and that a classifier trained by dataset from larger number of people does not always perform better. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.1109/GCCE46687.2019.9014644 | GCCE |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
0 | 2 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Trang Thuy Vu | 1 | 0 | 0.34 |
| Kaori Fujinami | 2 | 316 | 41.25 |