FEAST at Play: Feature ExtrAction using Score function Tensors

Majid Janzamin, Hanie Sedghi, U.N. Niranjan, Animashree Anandkumar
Proceedings of the 1st International Workshop on Feature Extraction: Modern Questions and Challenges at NIPS 2015, PMLR 44:130-144, 2015.

Abstract

Feature learning forms the cornerstone for tackling challenging learning problems in domains such as speech, computer vision and natural language processing. In this paper, we build upon a novel framework called FEAST(Feature ExtrAction using Score function Tensors) which incorporates generative models for discriminative learning. FEAST considers a novel class of matrix and tensor-valued feature transform, which can be pre-trained using unlabeled samples. It uses an efficient algorithm for extracting discriminative information, given these pre-trained features and labeled samples for any related task. The class of features it adopts are based on higher-order score functions, which capture local variations in the probability density function of the input. We employ efficient spectral decomposition algorithms (on matrices and tensors) for extracting discriminative components. The advantage of employing tensor-valued features is that we can extract richer discriminative information in the form of overcomplete representations (where number of discriminative features is greater than input dimension). In this paper, we provide preliminary experiment results on real datasets.

Cite this Paper

BibTeX
@InProceedings{pmlr-v44-janzamin2015, title = {{FEAST at Play: Feature ExtrAction using Score function Tensors}}, author = {Janzamin, Majid and Sedghi, Hanie and Niranjan, U.N. and Anandkumar, Animashree}, booktitle = {Proceedings of the 1st International Workshop on Feature Extraction: Modern Questions and Challenges at NIPS 2015}, pages = {130--144}, year = {2015}, editor = {Storcheus, Dmitry and Rostamizadeh, Afshin and Kumar, Sanjiv}, volume = {44}, series = {Proceedings of Machine Learning Research}, address = {Montreal, Canada}, month = {11 Dec}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v44/janzamin2015.pdf}, url = {https://proceedings.mlr.press/v44/janzamin2015.html}, abstract = {Feature learning forms the cornerstone for tackling challenging learning problems in domains such as speech, computer vision and natural language processing. In this paper, we build upon a novel framework called FEAST(Feature ExtrAction using Score function Tensors) which incorporates generative models for discriminative learning. FEAST considers a novel class of matrix and tensor-valued feature transform, which can be pre-trained using unlabeled samples. It uses an efficient algorithm for extracting discriminative information, given these pre-trained features and labeled samples for any related task. The class of features it adopts are based on higher-order score functions, which capture local variations in the probability density function of the input. We employ efficient spectral decomposition algorithms (on matrices and tensors) for extracting discriminative components. The advantage of employing tensor-valued features is that we can extract richer discriminative information in the form of overcomplete representations (where number of discriminative features is greater than input dimension). In this paper, we provide preliminary experiment results on real datasets.} }
Endnote
%0 Conference Paper %T FEAST at Play: Feature ExtrAction using Score function Tensors %A Majid Janzamin %A Hanie Sedghi %A U.N. Niranjan %A Animashree Anandkumar %B Proceedings of the 1st International Workshop on Feature Extraction: Modern Questions and Challenges at NIPS 2015 %C Proceedings of Machine Learning Research %D 2015 %E Dmitry Storcheus %E Afshin Rostamizadeh %E Sanjiv Kumar %F pmlr-v44-janzamin2015 %I PMLR %P 130--144 %U https://proceedings.mlr.press/v44/janzamin2015.html %V 44 %X Feature learning forms the cornerstone for tackling challenging learning problems in domains such as speech, computer vision and natural language processing. In this paper, we build upon a novel framework called FEAST(Feature ExtrAction using Score function Tensors) which incorporates generative models for discriminative learning. FEAST considers a novel class of matrix and tensor-valued feature transform, which can be pre-trained using unlabeled samples. It uses an efficient algorithm for extracting discriminative information, given these pre-trained features and labeled samples for any related task. The class of features it adopts are based on higher-order score functions, which capture local variations in the probability density function of the input. We employ efficient spectral decomposition algorithms (on matrices and tensors) for extracting discriminative components. The advantage of employing tensor-valued features is that we can extract richer discriminative information in the form of overcomplete representations (where number of discriminative features is greater than input dimension). In this paper, we provide preliminary experiment results on real datasets.
RIS
TY - CPAPER TI - FEAST at Play: Feature ExtrAction using Score function Tensors AU - Majid Janzamin AU - Hanie Sedghi AU - U.N. Niranjan AU - Animashree Anandkumar BT - Proceedings of the 1st International Workshop on Feature Extraction: Modern Questions and Challenges at NIPS 2015 DA - 2015/12/08 ED - Dmitry Storcheus ED - Afshin Rostamizadeh ED - Sanjiv Kumar ID - pmlr-v44-janzamin2015 PB - PMLR DP - Proceedings of Machine Learning Research VL - 44 SP - 130 EP - 144 L1 - http://proceedings.mlr.press/v44/janzamin2015.pdf UR - https://proceedings.mlr.press/v44/janzamin2015.html AB - Feature learning forms the cornerstone for tackling challenging learning problems in domains such as speech, computer vision and natural language processing. In this paper, we build upon a novel framework called FEAST(Feature ExtrAction using Score function Tensors) which incorporates generative models for discriminative learning. FEAST considers a novel class of matrix and tensor-valued feature transform, which can be pre-trained using unlabeled samples. It uses an efficient algorithm for extracting discriminative information, given these pre-trained features and labeled samples for any related task. The class of features it adopts are based on higher-order score functions, which capture local variations in the probability density function of the input. We employ efficient spectral decomposition algorithms (on matrices and tensors) for extracting discriminative components. The advantage of employing tensor-valued features is that we can extract richer discriminative information in the form of overcomplete representations (where number of discriminative features is greater than input dimension). In this paper, we provide preliminary experiment results on real datasets. ER -
APA
Janzamin, M., Sedghi, H., Niranjan, U. & Anandkumar, A.. (2015). FEAST at Play: Feature ExtrAction using Score function Tensors. Proceedings of the 1st International Workshop on Feature Extraction: Modern Questions and Challenges at NIPS 2015, in Proceedings of Machine Learning Research 44:130-144 Available from https://proceedings.mlr.press/v44/janzamin2015.html.

Related Material