Learning Multiple Related Tasks using Latent Independent Component Analysis
Download:
|
|
by Jian Zhangý, Zoubin Ghahramaniýþ, Yiming Yangý
http://www.cs.cmu.edu/~jianzhan/./papers/zgy-nips05.ps
Add To MetaCart
Abstract:
We propose a probabilistic model based on Independent Component Analysis for learning multiple related tasks. In our model the task parameters are assumed to be generated from independent sources which account for the relatedness of the tasks. We use Laplace distributions to model hidden sources which makes it possible to identify the hidden, independent components instead of just modeling correlations. Furthermore, our model enjoys a sparsity property which makes it both parsimonious and robust. We also propose efficient algorithms for both empirical Bayes method and point estimation. Our experimental results on two multi-label text classification data sets show that the proposed approach is promising. 1
Citations
| 289 | Hierarchically classifying documents using very few words – Koller, Sahami - 1997 |
| 92 | Variational inference for Bayesian mixtures of factor analysers – Ghahramani, Beal - 1999 |
| 47 | A model of inductive bias learning – Baxter |
| 28 | Predicting multivariate responses in multiple linear regression – Breiman, Friedman - 1997 |
| 22 | Learning multiple tasks with kernel methods – Evgeniou, Micchelli, et al. - 2005 |
| 16 | A variational approach to Bayesian logistic regression models and their extensions – Jaakkola, Jordan - 1997 |
| 11 | Empirical Bayes for learning to learn – Heskes - 2000 |
| 4 | Semiparametric Latent Factor Models – Seeger, Jordan, et al. - 2004 |
| 2 | A Framework for Learning Predicative Structures from Multiple Tasks and Unlabeled Data – Ando, Zhang - 2004 |
