Abstract
Feature embedding is an emerging research area which intends to transform features from the original space into a new space to support effective learning. Many feature embedding algorithms exist, but they often suffer from several major drawbacks, including (1) only handle single feature types, or users have to clearly separate features into different feature views and supply such information for feature embedding learning; (2) designed for either supervised or unsupervised learning tasks, but not for both; and (3) feature embedding for new out-of-training samples have to be obtained through a retraining phase, therefore unsuitable for online learning tasks. In this paper, we propose a generalized feature embedding algorithm, GEL, for both supervised, unsupervised, and online learning tasks. GEL learns feature embedding from any type of data or data with mixed feature types. For supervised learning tasks with class label information, GEL leverages a Class Partitioned Instance Representation (CPIR) process to arrange instances, based on their labels, as a dense binary representation via row and feature vectors for feature embedding learning. If class labels are unavailable, CPIR is naturally degenerated and treats all instances as one class. Based on the CPIR representation, GEL uses eigenvector decomposition to convert the proximity matrix into a low-dimensional space. For new out-of-training samples, their low-dimensional representation are derived through a direct conversion without a retraining phase. The learned numerical embedding features can be directly used to represent instances for effective learning. Experiments and comparisons on 28 datasets, including categorical, numerical, and ordinal features, demonstrate that embedding features learned from GEL can effectively represent the original instances for clustering, classification, and online learning.
Similar content being viewed by others
References
Abdi, H., & Valentin, D. (2007). Multiple correspondence analysis. In Encyclopedia of measurement and statistics (pp. 651–657).
Alamuri, M., Surampudi, B.R., Negi, A. (2014). A survey of distance/similarity measures for categorical data. In 2014 International joint conference on neural networks (IJCNN) (pp. 1907–1914).
Aljarah, I. (2016). https://www.kaggle.com/aljarah/xapi-edu-data.
Argyriou, A., & Evgeniou, T. (2007). Multi-task feature learning. In Proceedings of neural information processing systems (NIPS).
Argyriou, A., Evgeniou, T., Pontil, M. (2008). Convex multi-task feature learning. Machine Learning, 73(3), 243–272.
Axler, S.J. (1997). Linear algebra done right Vol. 2. Berlin: Springer.
Babenko, A., & Lempitsky, V. (2015). Aggregating local deep features for image retrieval. In Proceedings of the IEEE international conference on computer vision (pp. 1269–1277).
Bates, D., & Eddelbuettel, D. (2013). Fast and elegant numerical linear algebra using the RcppEigen package. Journal of Statistical Software, 52(5), 1–24.
Benoit, K., & Nulty, P. (2016). quanteda: quantitative analysis of textual data. R package version 0.9, 8.
Bro, R., & Smilde, A.K. (2014). Principal component analysis. Analytical Methods, 6(9), 2812–2831.
Chen, C., Shyu, M.-L., Chen, S.-C. (2016). Weighted subspace modeling for semantic concept retrieval using gaussian mixture models. Information Systems Frontiers, 18(5), 877–889.
Choi, S.-S., Cha, S.-H., Tappert, C.C. (2010). A survey of binary similarity and distance measures. Journal of Systemics, Cybernetics and Informatics, 8(1), 43–48.
Crane, H. (2015). Clustering from categorical data sequences. Journal of the American Statistical Association, 110(510), 810–823.
de Leeuw, J. (2011). Principal component analysis of binary data. applications to roll-call-analysis. Department of statistics, UCLA.
Ditzler, G., & Polikar, R. (2013). Incremental learning of concept drift from streaming imbalanced data. ieee transactions on knowledge and data engineering, 25(10), 2283–2301.
Domingos, P. (2012). A few useful things to know about machine learning. Communications of the ACM, 55 (10), 78–87.
Gal, Y., Chen, Y., Ghahramani, Z. (2015). Latent gaussian processes for distribution estimation of multivariate categorical data. In Proceedings of the 32nd international conference on machine learning (ICML-15) (pp. 645–654).
Gelbard, R. (2013). padding bitmaps to support similarity and mining. Information Systems Frontiers, 15(1), 99–110.
Golinko, E., & Zhu, X. (2017). Gfel: Generalized feature embedding learning using weighted instance matching. In 2017 IEEE International conference on information reuse and integration (IRI) (pp. 235–244).
Greenacre, M. (2007). Correspondence analysis in practice. CRC press.
Greene, D. (2016). http://mlg.ucd.ie/datasets/bbc.html.
Guyon, I., & Elisseeff, A. (2003). An introduction to variable and feature selection. Journal of machine learning research 3:1157–1182.
Hou, C., Nie, F., Li, X., Yi, D., Wu, Y. (2014). Joint embedding learning and sparse regression: A framework for unsupervised feature selection. IEEE Transactions on Cybernetics, 44(6), 793–804.
Hsu, C.-W., Chang, C.-C., Lin, C.-J., et al. (2003). A practical guide to support vector classification.
Hsu, C.-C., & Huang, W.-H. (2016). Integrated dimensionality reduction technique for mixed-type data involving categorical values. Applied Soft Computing, 43, 199–209.
Jia, Y., Shelhamer, E., Donahue, J., Karayev, S., Long, J., Girshick, R., Guadarrama, S., Darrell, T. (2014). Caffe: Convolutional architecture for fast feature embedding. In Proceedings of ACM multimedia conference.
Juan, A., & Vidal, E. (2004). Bernoulli mixture models for binary images. In Proceedings of the 17th international conference on Pattern recognition, 2004. ICPR 2004, (Vol. 3 pp. 367–370). IEEE.
Kaban, A., Bingham, E., Hirsimäki, T. (2004). Learning to read between the lines The aspect bernoulli model. In Proceedings of the 2004 SIAM international conference on data mining (pp. 462–466). SIAM.
Kaggle. (2017). https://www.kaggle.com.
Krijthe, J. (2015). Rtsne: T-distributed stochastic neighbor embedding using barnes-hut implementation. R package version 0.10, http://CRAN.R-project.org/package=Rtsne.
Lee, S. (2009). Principal components analysis for binary data. PhD thesis: Texas A&M University.
Lee, S., Huang, J.Z., Hu, J. (2010). Sparse logistic principal components analysis for binary data. The annals of applied statistics, 4(3), 1579.
Lichman, M. (2013). UCI machine learning repository.
van der Maaten, L., & Hinton, G. (2008). Visualizing data using t-sne. Journal of Machine Learning Research, 9(Nov), 2579–2605.
Malik, Z.K., Hussain, A., Wu, J. (2016). An online generalized eigenvalue version of laplacian eigenmaps for visual big data. Neurocomputing, 173, 127–136.
Meyer, D., & Buchta, C. proxy: Distance and Similarity Measures, 2016. R package version 0.4-16.
Muhlbaier, M.D., & Polikar, R. (2007). An ensemble approach for incremental learning in nonstationary environments. In International workshop on multiple classifier systems (pp. 490–500). Berlin: Springer.
Müller, B., Reinhardt, J., Strickland, M.T. (2012). Neural networks: an introduction. Berlin: Springer Science & Business Media.
Najafi, A., Motahari, A., Rabiee, H.R. (2017). Reliable learning of bernoulli mixture models. arXiv:1710.02101.
Nenadic, O., & Greenacre, M. (2007). Correspondence analysis in r, with two-and three-dimensional graphics The ca package. Journal of Statistical Software.
Pan, S., Wu, J.W., Zhu, X., Zhang, C., Wang, Y. (2016). Tri-party deep network representation. In Proc. of international joint conference on artificial intelligence.
Plaza, A., Benediktsson, J.A., Boardman, J.W., Brazile, J., Bruzzone, L., Camps-Valls, G., Chanussot, J., Fauvel, M., Gamba, P., Gualtieri, A., et al. (2009). Recent advances in techniques for hyperspectral image processing. Remote sensing of environment, 113, S110–S122.
Qian, Y., Li, F., Liang, J., Liu, B., Dang, C. (2016). Space structure and clustering of categorical data. IEEE transactions on neural networks and learning systems, 27(10), 2047–2059.
Ramos, J. (2003). Using tf-idf to determine word relevance in document queries. In Proceedings of the first instructional conference on machine learning.
Rokach, L., & Maimon, O. (2005). Decision trees. Data mining and knowledge discovery handbook, pp. 165–192.
Rokach, L., & Maimon, O. (2014). Data mining with decision trees: theory and applications. Singapore: World scientific.
Romero, C., Ventura, S., Espejo, P.G., Hervás, C. (2008). Data mining algorithms to classify students. In Educational data mining 2008.
Roweis, S., & Saul, L. (2000). Nonlinear dimensionality reduction by locally linear embedding. Science, 290 (5500), 2323–2326.
Shen, L., Wang, H., Xu, L.D., Ma, X., Chaudhry, S., He, W. (2016). Identity management based on pca and svm. Information Systems Frontiers, 18(4), 711–716.
Shlens, J. (2014). A tutorial on principal component analysis. arXiv:1404.1100.
Shmelkov, K., Schmid, C., Alahari, K. (2017). Incremental learning of object detectors without catastrophic forgetting. arXiv:1708.06977.
Stanford. (2009). https://nlp.stanford.edu/ir-book/html/htmledition/evaluation-of-clustering-1.html.
Strange, H., & Zwiggelaar, R. (2011). A generalised solution to the out-of-sample extension problem in manifold learning. In AAAI (pp. 293–296).
Sun, B.-Y., Zhang, X.-M., Li, J., Mao, X.-M. (2010). Feature fusion using locally linear embedding for classification. IEEE Transactions on Neural Networks, 21(1), 163–168.
Tsymbal, A., Puuronen, S., Pechenizkiy, M., Baumgarten, M., Patterson, D.W. (2002). Eigenvector-based feature extraction for classification. In FLAIRS Conference (pp. 354–358).
Venables, W.N., & Ripley, B.D. (2002). Modern applied statistics with S, 4th edn. New York: Springer. ISBN 0-387-95457-0.
Vural, E., & Guillemot, C. (2016). Out-of-sample generalizations for supervised manifold learning for classification. IEEE Transactions on Image Processing, 25(3), 1410–1424.
Xie, J., Szymanski, B.K., Zaki, M.J. (2010). Learning dissimilarities for categorical symbols. FSDM, 10, 97–106.
Zhang, D., Yin, J., Zhu, X., Zhang, C. (2017). User profile preserving social network embedding. In Proc. of international joint conference on artificial intelligence.
Zhang, H. (2004). The optimality of naive bayes. AA, 1(2), 3.
Zhang, L., Zhang, Q., Zhang, L., Tao, D., Huang, X., Du, B. (2015). Ensemble manifold regularized sparse low-rank approximation for multiview feature embedding. Pattern Recognition, 48(10).
Zhang, P., Zhu, X., Shi, Y. (2008). Categorizing and mining concept drifting data streams. In ACM SIGKDD Conference (pp. 812–820).
Zheng, L., Wang, S., Tian, Q. (2014). Coupled binary embedding for large-scale image retrieval. IEEE Transactions on Image processing, 23(8), 3368–3380.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix
Toy example showing detailed GEL feature embedding learning process
Converting to binary
Then we compute the weighted proximity matrix
Then
And,
Then
Then
Finally, we utilize single value decomposition the find the first two eigenvectors V2 of,
And we reduce dimension only on the first two components
and
Link to code
We hereby release the source code for public examination and validation: https://github.com/egolinko/GEL
Rights and permissions
About this article
Cite this article
Golinko, E., Zhu, X. Generalized Feature Embedding for Supervised, Unsupervised, and Online Learning Tasks. Inf Syst Front 21, 125–142 (2019). https://doi.org/10.1007/s10796-018-9850-y
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10796-018-9850-y