Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 47615 (19)
- University: Sharif University of Technology
- Department: Computer Engineering
- Advisor(s): Soleymani Baghshah, Mahdieh
- Abstract:
- A predefined assumption in many learning algorithms is that the training and test data must be in thesame feature space and have the same distribution.However, this assumption may not hold in all of these algorithms and in the real world there might be difference between the source and the targer domian, whether in the feature space or the distribution. Moreover, there might be a few number of labled data of the target domain which causes difficulty in learning an accurate classifier. In such cases, transferring knowledge can be useful if can be done successfully and transfer learning was introduced for this purpose. Domain Adaptation is one of the transfer leaning problems that assume some conditions on the feature space and the distribution of data in the source and the target domain. One of the most successful approaches in domain adaptation is respresentation learnig which tries to find a new represention space in which soure and target data can be represented such that learnig a classifier with the source domain data leads also to a good classification for the target domain data. We use deep learning approach that is a powerful representation learning method and introduce three approaches to address problems of the existing methods. The results show, somehow, we achieve to better and faster algorithms
- Keywords:
- Domain Adaptation ; Deep Learning ; Representation Learning
Digital Object List-
محتواي کتاب - view
Bookmark