Multi-label classification via label correlation and first order feature dependence features in a data stream.
Nguyen, Tien Thanh
Nguyen, Thi Thu Thuy
Luong, Anh Vu
Nguyen, Quoc Viet Hung
Liew, Alan Wee-Chung
MetadataShow full item record
NGUYEN, T.T., NGUYEN, T.T.T., LUONG, A.V., NGUYEN, Q.V.H., LIEW, A.W.-C. and STANTIC, B. 2019. Multi-label classification via label correlation and first order feature dependance in a data stream. Pattern recognition [online], 90, pages 35-51. Available from: https://doi.org/10.1016/j.patcog.2019.01.007
Many batch learning algorithms have been introduced for offline multi-label classification (MLC) over the years. However, the increasing data volume in many applications such as social networks, sensor networks, and traffic monitoring has posed many challenges to batch MLC learning. For example, it is often expensive to re-train the model with the newly arrived samples, or it is impractical to learn on the large volume of data at once. The research on incremental learning is therefore applicable to a large volume of data and especially for data stream. In this study, we develop a Bayesian-based method for learning from multi-label data streams by taking into consideration the correlation between pairs of labels and the relationship between label and feature. In our model, not only the label correlation is learned with each arrived sample with ground truth labels but also the number of predicted labels are adjusted based on Hoeffding inequality and the label cardinality. We also extend the model to handle missing values, a problem common in many real-world data. To handle concept drift, we propose a decay mechanism focusing on the age of the arrived samples to incrementally adapt to the change of data. The experimental results show that our method is highly competitive compared to several well-known benchmark algorithms under both the stationary and concept drift settings. Please note that the published title differs from this accepted manuscript "Multi-label classification via labels correlation and one-dependence features on data stream."
Permalink for this recordhttp://hdl.handle.net/10059/3301
Collections in which this item appears
Except where otherwise noted, this item's license is described as https://creativecommons.org/licenses/by-nc-nd/4.0
Showing items related by title, author, creator and subject.
Grinnall, Andy; Burnett, Simon M. (MDPI http://dx.doi.org/10.3390/su7056086, 2015-05)GRINNALL, A. and BURNETT, S., 2015. First catch your fish: designing a “low energy fish” label. Sustainability, 7 (5), pp. 6086-6101.This paper explores the application of information design principles to label design for fish packaging, identifying energy implications for the product. This stage of the project has consisted of: A review and distillation ...
Oates, Caroline J.; Alevizou, Panayiota J.; McDonald, Seonaidh (MDPI https://doi.org/10.3390/su8010075, 2016-01-13)OATES, C., ALEVIZOU, P. and MCDONALD, S. 2016. Challenges for marketers in sustainable production and consumption. Sustainability [online], 8(1), article 75. Available from: https://doi.org/10.3390/su8010075.As one of the biggest issues facing today's global society, sustainability cuts across all areas of production and consumption and presents challenges for marketers who attempt to understand and incorporate sustainability ...
Alevizou, Panayiota J.; Oates, Caroline J.; McDonald, Seonaidh (MDPI http://dx.doi.org/10.3390/su7078729, 2015-07)ALEVIZOU, P. J., OATES, C. J. and MCDONALD, S., 2015. The well(s) of knowledge: the decoding of sustainability claims in the UK and in Greece. Sustainability, 7 (7), pp. 8729-8747.Sustainability claims have existed on fast moving consumer goods (FMCGs) for over four decades and there is evidence that they are increasing. Research suggests that consumers have a low level of knowledge and understanding ...