Journal
ACM TRANSACTIONS ON KNOWLEDGE DISCOVERY FROM DATA
Volume 13, Issue 2, Pages -Publisher
ASSOC COMPUTING MACHINERY
DOI: 10.1145/3309541
Keywords
Feature selection; probabilistic classification model; sparse Bayesian learning; supervised learning; EEG emotion recognition
Funding
- National Natural Science Foundation of China [91846111, 91746209]
- Science and Technology Innovation Committee Foundation of Shenzhen [ZDSYS201703031748284]
- Ahold Delhaize
- Amsterdam Data Science
- Bloomberg Research Grant program
- China Scholarship Council
- Criteo Faculty Research Award program
- Elsevier
- European Community [312827]
- Google Faculty Research Awards program
- Microsoft Research Ph.D. program
- Netherlands Institute for Sound and Vision
- Netherlands Organisation for Scientific Research (NWO) [CI-14-25, 652.002.001, 612.001.551, 652.001.003]
- Yandex
Ask authors/readers for more resources
Sparse Bayesian learning is a state-of-the-art supervised learning algorithm that can choose a subset of relevant samples from the input data and make reliable probabilistic predictions. However, in the presence of high-dimensional data with irrelevant features, traditional sparse Bayesian classifiers suffer from performance degradation and low efficiency due to the incapability of eliminating irrelevant features. To tackle this problem, we propose a novel sparse Bayesian embedded feature selection algorithm that adopts truncated Gaussian distributions as both sample and feature priors. The proposed algorithm, called probabilistic feature selection and classification vector machine (PFCVMLP) is able to simultaneously select relevant features and samples for classification tasks. In order to derive the analytical solutions, Laplace approximation is applied to compute approximate posteriors and marginal likelihoods. Finally, parameters and hyperparameters are optimized by the type-II maximum likelihood method. Experiments on three datasets validate the performance of PFCVMLP along two dimensions: classification performance and effectiveness for feature selection. Finally, we analyze the generalization performance and derive a generalization error bound for PFCVMLP. By tightening the bound, the importance of feature selection is demonstrated.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available