Discriminative Embeddings of Latent Variable Models for Structured Data

Hanjun Dai, Bo Dai, Le Song
; Proceedings of The 33rd International Conference on Machine Learning, PMLR 48:2702-2711, 2016.

Abstract

Kernel classifiers and regressors designed for structured data, such as sequences, trees and graphs, have significantly advanced a number of interdisciplinary areas such as computational biology and drug design. Typically, kernels are designed beforehand for a data type which either exploit statistics of the structures or make use of probabilistic generative models, and then a discriminative classifier is learned based on the kernels via convex optimization. However, such an elegant two-stage approach also limited kernel methods from scaling up to millions of data points, and exploiting discriminative information to learn feature representations. We propose, structure2vec, an effective and scalable approach for structured data representation based on the idea of embedding latent variable models into feature spaces, and learning such feature spaces using discriminative information. Interestingly, structure2vec extracts features by performing a sequence of function mappings in a way similar to graphical model inference procedures, such as mean field and belief propagation. In applications involving millions of data points, we showed that structure2vec runs 2 times faster, produces models which are 10,000 times smaller, while at the same time achieving the state-of-the-art predictive performance.

Cite this Paper


BibTeX
@InProceedings{pmlr-v48-daib16, title = {Discriminative Embeddings of Latent Variable Models for Structured Data}, author = {Hanjun Dai and Bo Dai and Le Song}, booktitle = {Proceedings of The 33rd International Conference on Machine Learning}, pages = {2702--2711}, year = {2016}, editor = {Maria Florina Balcan and Kilian Q. Weinberger}, volume = {48}, series = {Proceedings of Machine Learning Research}, address = {New York, New York, USA}, month = {20--22 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v48/daib16.pdf}, url = {http://proceedings.mlr.press/v48/daib16.html}, abstract = {Kernel classifiers and regressors designed for structured data, such as sequences, trees and graphs, have significantly advanced a number of interdisciplinary areas such as computational biology and drug design. Typically, kernels are designed beforehand for a data type which either exploit statistics of the structures or make use of probabilistic generative models, and then a discriminative classifier is learned based on the kernels via convex optimization. However, such an elegant two-stage approach also limited kernel methods from scaling up to millions of data points, and exploiting discriminative information to learn feature representations. We propose, structure2vec, an effective and scalable approach for structured data representation based on the idea of embedding latent variable models into feature spaces, and learning such feature spaces using discriminative information. Interestingly, structure2vec extracts features by performing a sequence of function mappings in a way similar to graphical model inference procedures, such as mean field and belief propagation. In applications involving millions of data points, we showed that structure2vec runs 2 times faster, produces models which are 10,000 times smaller, while at the same time achieving the state-of-the-art predictive performance.} }
Endnote
%0 Conference Paper %T Discriminative Embeddings of Latent Variable Models for Structured Data %A Hanjun Dai %A Bo Dai %A Le Song %B Proceedings of The 33rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2016 %E Maria Florina Balcan %E Kilian Q. Weinberger %F pmlr-v48-daib16 %I PMLR %J Proceedings of Machine Learning Research %P 2702--2711 %U http://proceedings.mlr.press %V 48 %W PMLR %X Kernel classifiers and regressors designed for structured data, such as sequences, trees and graphs, have significantly advanced a number of interdisciplinary areas such as computational biology and drug design. Typically, kernels are designed beforehand for a data type which either exploit statistics of the structures or make use of probabilistic generative models, and then a discriminative classifier is learned based on the kernels via convex optimization. However, such an elegant two-stage approach also limited kernel methods from scaling up to millions of data points, and exploiting discriminative information to learn feature representations. We propose, structure2vec, an effective and scalable approach for structured data representation based on the idea of embedding latent variable models into feature spaces, and learning such feature spaces using discriminative information. Interestingly, structure2vec extracts features by performing a sequence of function mappings in a way similar to graphical model inference procedures, such as mean field and belief propagation. In applications involving millions of data points, we showed that structure2vec runs 2 times faster, produces models which are 10,000 times smaller, while at the same time achieving the state-of-the-art predictive performance.
RIS
TY - CPAPER TI - Discriminative Embeddings of Latent Variable Models for Structured Data AU - Hanjun Dai AU - Bo Dai AU - Le Song BT - Proceedings of The 33rd International Conference on Machine Learning PY - 2016/06/11 DA - 2016/06/11 ED - Maria Florina Balcan ED - Kilian Q. Weinberger ID - pmlr-v48-daib16 PB - PMLR SP - 2702 DP - PMLR EP - 2711 L1 - http://proceedings.mlr.press/v48/daib16.pdf UR - http://proceedings.mlr.press/v48/daib16.html AB - Kernel classifiers and regressors designed for structured data, such as sequences, trees and graphs, have significantly advanced a number of interdisciplinary areas such as computational biology and drug design. Typically, kernels are designed beforehand for a data type which either exploit statistics of the structures or make use of probabilistic generative models, and then a discriminative classifier is learned based on the kernels via convex optimization. However, such an elegant two-stage approach also limited kernel methods from scaling up to millions of data points, and exploiting discriminative information to learn feature representations. We propose, structure2vec, an effective and scalable approach for structured data representation based on the idea of embedding latent variable models into feature spaces, and learning such feature spaces using discriminative information. Interestingly, structure2vec extracts features by performing a sequence of function mappings in a way similar to graphical model inference procedures, such as mean field and belief propagation. In applications involving millions of data points, we showed that structure2vec runs 2 times faster, produces models which are 10,000 times smaller, while at the same time achieving the state-of-the-art predictive performance. ER -
APA
Dai, H., Dai, B. & Song, L.. (2016). Discriminative Embeddings of Latent Variable Models for Structured Data. Proceedings of The 33rd International Conference on Machine Learning, in PMLR 48:2702-2711

Related Material