TY - GEN
T1 - Differential mutual information forward search for multi-kernel discriminant-component selection with an application to privacy-preserving classification
AU - Chanyaswad, Thee
AU - Al, Mert
AU - Chang, J. Morris
AU - Kung, S. Y.
N1 - Publisher Copyright:
© 2017 IEEE.
PY - 2017/12/5
Y1 - 2017/12/5
N2 - In machine learning, feature engineering has been a pivotal stage in building a high-quality predictor. Particularly, this work explores the multiple Kernel Discriminant Component Analysis (mKDCA) feature-map and its variants. However, seeking the right subset of kernels for mKDCA feature-map can be challenging. Therefore, we consider the problem of kernel selection, and propose an algorithm based on Differential Mutual Information (DMI) and incremental forward search. DMI serves as an effective metric for selecting kernels, as is theoretically supported by mutual information and Fisher's discriminant analysis. On the other hand, incremental forward search plays a role in removing redundancy among kernels. Finally, we illustrate the potential of the method via an application in privacy-aware classification, and show on three mobile-sensing datasets that selecting an effective set of kernels for mKDCA feature-maps can enhance the utility classification performance, while successfully preserve the data privacy. Specifically, the results show that the proposed DMI forward search method can perform better than the state-of-the-art, and, with much smaller computational cost, can perform as well as the optimal, yet computationally expensive, exhaustive search.
AB - In machine learning, feature engineering has been a pivotal stage in building a high-quality predictor. Particularly, this work explores the multiple Kernel Discriminant Component Analysis (mKDCA) feature-map and its variants. However, seeking the right subset of kernels for mKDCA feature-map can be challenging. Therefore, we consider the problem of kernel selection, and propose an algorithm based on Differential Mutual Information (DMI) and incremental forward search. DMI serves as an effective metric for selecting kernels, as is theoretically supported by mutual information and Fisher's discriminant analysis. On the other hand, incremental forward search plays a role in removing redundancy among kernels. Finally, we illustrate the potential of the method via an application in privacy-aware classification, and show on three mobile-sensing datasets that selecting an effective set of kernels for mKDCA feature-maps can enhance the utility classification performance, while successfully preserve the data privacy. Specifically, the results show that the proposed DMI forward search method can perform better than the state-of-the-art, and, with much smaller computational cost, can perform as well as the optimal, yet computationally expensive, exhaustive search.
KW - Compressive Privacy
KW - Differential Mutual Information (DMI)
KW - Incremental forward search
KW - Kernel Discriminant Component Analysis (KDCA)
KW - Kernel selection
KW - Multi-kernel learning
UR - http://www.scopus.com/inward/record.url?scp=85042312591&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85042312591&partnerID=8YFLogxK
U2 - 10.1109/MLSP.2017.8168177
DO - 10.1109/MLSP.2017.8168177
M3 - Conference contribution
AN - SCOPUS:85042312591
T3 - IEEE International Workshop on Machine Learning for Signal Processing, MLSP
SP - 1
EP - 6
BT - 2017 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2017 - Proceedings
A2 - Ueda, Naonori
A2 - Chien, Jen-Tzung
A2 - Matsui, Tomoko
A2 - Larsen, Jan
A2 - Watanabe, Shinji
PB - IEEE Computer Society
T2 - 2017 IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2017
Y2 - 25 September 2017 through 28 September 2017
ER -