Multiple-Instance Learning via Random Walk
Abstract
This paper presents a decoupled two stage solution to the multiple-instance learning (MIL) problem. With a constructed affinity matrix to reflect the instance relations, a modified Random Walk on a Graph process is applied to infer the positive instances in each positive bag. This process has both a closed form solution and an efficient iterative one. Combined with the Support Vector Machine (SVM) classifier, this algorithm decouples the inferring and training stages and converts MIL into a supervised learning problem. Compared with previous algorithms on several benchmark data sets, the proposed algorithm is quite competitive in both computational efficiency and classification accuracy.
Cite
Text
Wang et al. "Multiple-Instance Learning via Random Walk." European Conference on Machine Learning, 2006. doi:10.1007/11871842_45Markdown
[Wang et al. "Multiple-Instance Learning via Random Walk." European Conference on Machine Learning, 2006.](https://mlanthology.org/ecmlpkdd/2006/wang2006ecml-multipleinstance/) doi:10.1007/11871842_45BibTeX
@inproceedings{wang2006ecml-multipleinstance,
title = {{Multiple-Instance Learning via Random Walk}},
author = {Wang, Dong and Li, Jianmin and Zhang, Bo},
booktitle = {European Conference on Machine Learning},
year = {2006},
pages = {473-484},
doi = {10.1007/11871842_45},
url = {https://mlanthology.org/ecmlpkdd/2006/wang2006ecml-multipleinstance/}
}