Knowl Inf Syst (2017) 52:687–707
Online transfer learning by leveraging multiple source
· Xiaoming Zhou
· Yuguang Yan
· Huaqing Min
Received: 5 April 2016 / Accepted: 28 December 2016 / Published online: 11 January 2017
© Springer-Verlag London 2017
Abstract Transfer learning aims to enhance performance in a target domain by exploiting
useful information from auxiliary or source domains when the labeled data in the target
domain are insufﬁcient or difﬁcult to acquire. In some real-world applications, the data
of source domain are provided in advance, but the data of target domain may arrive in
a stream fashion. This kind of problem is known as online transfer learning. In practice,
there can be several source domains that are related to the target domain. The performance
of online transfer learning is highly associated with selected source domains, and simply
combining the source domains may lead to unsatisfactory performance. In this paper, we
seek to promote classiﬁcation performance in a target domain by leveraging labeled data
from multiple source domains in online setting. To achieve this, we propose a new online
transfer learning algorithm that merges and leverages the classiﬁers of the source and target
domain with an ensemble method. The mistake bound of the proposed algorithm is analyzed,
and the comprehensive experiments on three real-world data sets illustrate that our algorithm
outperforms the compared baseline algorithms.
Keywords Transfer learning · Online learning · Online transfer learning · multiple source
In machine learning technologies, most models are built from large quantities of training data.
However, it is usually expensive and time-consuming to collect and label the training data.
Qingyao Wu, Xiaoming Zhou: Co-ﬁrst author.
School of Software Engineering, South China University of Technology, Guangzhou, China