Domain adaptation on graphs by learning graph topologies: theoretical analysis and an algorithm

2019-01-01
Vural, Elif
Show full item record
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
259
views
0
downloads
Traditional machine learning algorithms assume that the training and test data have the same distribution, while this assumption does not necessarily hold in real applications. Domain adaptation methods take into account the deviations in data distribution. In this work, we study the problem of domain adaptation on graphs. We consider a source graph and a target graph constructed with samples drawn from data manifolds. We study the problem of estimating the unknown class labels on the target graph using the label information on the source graph and the similarity between the two graphs. We particularly focus on a setting where the target label function is learned such that its spectrum is similar to that of the source label function. We first propose a theoretical analysis of domain adaptation on graphs and present performance bounds that characterize the target classification error in terms of the properties of the graphs and the data manifolds. We show that the classification performance improves as the topologies of the graphs get more balanced, i.e. as the numbers of neighbors of different graph nodes become more proportionate, and weak edges with small weights are avoided. Our results also suggest that graph edges between too distant data samples should be avoided for good generalization performance. We then propose a graph domain adaptation algorithm inspired by our theoretical findings, which estimates the label functions while learning the source and target graph topologies at the same time. The joint graph learning and label estimation problem is formulated through an objective function relying on our performance bounds, which is minimized with an alternating optimization scheme. Experiments on synthetic and real data sets suggest that the proposed method outperforms baseline approaches.
Electrical and Electronic Engineering, General Computer Science
https://hdl.handle.net/11511/47534
TURKISH JOURNAL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCES
Department of Electrical and Electronics Engineering, Article

Suggestions

Direction finding accuracy of sequential lobing under target amplitude fluctuations
Candan, Çağatay (Institution of Engineering and Technology (IET), 2015-01-01)
Using recently developed statistical target fluctuation models, the accuracy of sequential lobing is analytically studied. The study shows that the sequential lobing method suffers from a significant performance loss, in comparison with the monopulse method, for the Rayleigh fluctuation model. For other fluctuation models, the performance loss gradually decreases as the amplitude spread associated with the fluctuation gets smaller. The present study aims to analytically quantify the mentioned accuracy loss ...
Schedulability analysis of real-time multiframe cosimulations on multicore platforms
Ahsan, Muhammad Uzair; Oğuztüzün, Mehmet Halit S. (The Scientific and Technological Research Council of Turkey, 2019-01-01)
For real-time simulations, the fidelity of simulation depends not only on the functional accuracy of simulation but also on its timeliness. It is helpful for simulation designers if they can analyze and verify that a simulation will always meet its timing requirements without unnecessarily sacrificing functional accuracy. Abstracting the simulated processes simply as software tasks allows us to transform the problem of verifying timeliness into a schedulability analysis problem where tasks are checked as to...
Domain Adaptation on Graphs via Frequency Analysis
Pilancı, Mehmet; Vural, Elif (2019-08-22)
Classical machine learning algorithms assume the training and test data to be sampled from the same distribution, while this assumption may be violated in practice. Domain adaptation methods aim to exploit the information available in a source domain in order to improve the performance of classification in a target domain. In this work, we focus on the problem of domain adaptation in graph settings. We consider a source graph with many labeled nodes and aim to estimate the class labels on a target graph wit...
A pattern classification approach for boosting with genetic algorithms
Yalabık, Ismet; Yarman Vural, Fatoş Tunay; Üçoluk, Göktürk; Şehitoğlu, Onur Tolga (2007-11-09)
Ensemble learning is a multiple-classifier machine learning approach which produces collections and ensembles statistical classifiers to build up more accurate classifier than the individual classifiers. Bagging, boosting and voting methods are the basic examples of ensemble learning. In this study, a novel boosting technique targeting to solve partial problems of AdaBoost, a well-known boosting algorithm, is proposed. The proposed system finds an elegant way of boosting a bunch of classifiers successively ...
Random Set Methods Estimation of Multiple Extended Objects
Granstrom, Karl; Lundquist, Christian; Gustafsson, Fredrik; Orguner, Umut (Institute of Electrical and Electronics Engineers (IEEE), 2014-06-01)
Random set-based methods have provided a rigorous Bayesian framework and have been used extensively in the last decade for point object estimation. In this article, we emphasize that the same methodology offers an equally powerful approach to estimation of so-called extended objects, i.e., objects that result in multiple detections on the sensor side. Building upon the analogy between Bayesian state estimation of a single object and random finite set (RFS) estimation for multiple objects, we give a tutorial...
Citation Formats
E. Vural, “Domain adaptation on graphs by learning graph topologies: theoretical analysis and an algorithm,” TURKISH JOURNAL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCES, pp. 1619–1635, 2019, Accessed: 00, 2020. [Online]. Available: https://hdl.handle.net/11511/47534.
METU IIS - Integrated Information Service open@metu.edu.tr