Welcome to the ICML Discussion Site!
The 2010 Discussion Site is hosted at http://mldiscuss.appspot.com
The aim of this site is to enable the research community to discuss papers accepted to ICML. Each paper accepted to ICML 2009 has a page on this site. Each page has details of the paper and a discussion thread.
More information about how to use this site – including how to include LaTeX expressions in the discussions – can be found in the Help section.
Also, bear in mind that this site is an experiment - if you have any problems, questions or suggestions you can leave Feedback.
Register before commenting: At present, if you leave comments without registering as a user of this site you will not be able to edit or delete them once they are posted.
Here are most recently updated discussions for ICML 2009 papers:
|Supervised Learning from Multiple Experts: Whom to trust when everyone lies a bit (2009)||2011/02/07 19:08||4 Comments|
|Learning Nonlinear Dynamic Models (2009)||2011/02/07 19:08||5 Comments|
|Structure Preserving Embedding (2009)||2010/05/13 08:32||1 Comment|
|Multi-Assignment Clustering for Boolean Data (2009)||2009/10/08 06:55||1 Comment|
|Stochastic Methods for L1 Regularized Loss Minimization (2009)||2009/08/04 17:12||1 Comment|
|Discriminative $k$ metrics (2009)||2009/08/02 04:39||1 Comment|
|Online Dictionary Learning for Sparse Coding (2009)||2009/07/09 13:02||2 Comments|
|A majorization-minimization algorithm for (multiple) hyperparameter learning (2009)||2009/07/06 07:26||2 Comments|
|Near-Bayesian Exploration in Polynomial Time (2009)||2009/06/30 01:52||7 Comments|
|Function factorization using warped Gaussian processes (2009)||2009/06/24 18:50||1 Comment|
|Regularization and Feature Selection in Least-Squares Temporal Difference Learning (2009)||2009/06/08 15:57||3 Comments|
|Information Theoretic Measures for Clusterings Comparison: Is a Correction for Chance Necessary? (2009)||2009/06/01 17:01||2 Comments|
|Surrogate Regret Bounds for Proper Losses (2009)||2009/05/25 01:14||1 Comment|
Here are the most recently updated discussions for the ICML 2008 papers:
|A Least Squares Formulation for Canonical Correlation Analysis (2008)||2013/01/14 20:33||2 Comments|
|Self-taught Clustering (2008)||2012/07/11 04:08||3 Comments|
|The GroupLASSO for Generalized Linear Models: uniqueness of solutions and efficient algorithms (2008)||2012/03/30 04:56||1 Comment|
|An HDP-HMM for Systems with State Persistence (2008)||2012/03/20 13:20||8 Comments|
|SVM Optimization: Inverse Dependence on Training Set Size (2008)||2011/02/07 19:07||3 Comments|
|Hierarchical sampling for active learning (2008)||2009/06/17 19:49||3 Comments|
|Modeling Interleaved Hidden Processes (2008)||2009/02/18 12:03||4 Comments|
|Prediction with expert advice for the Brier game (2008)||2008/10/28 14:59||1 Comment|
|Efficient Bandit Algorithms for Online Multiclass Prediction (2008)||2008/07/26 13:45||1 Comment|
|On the Quantitative Analysis of Deep Belief Networks (2008)||2008/07/25 15:40||3 Comments|
|Knows What It Knows: A Framework For Self-Aware Learning (2008)||2008/07/12 03:50||7 Comments|
|Estimating Labels from Label Proportions (2008)||2008/07/09 01:15||1 Comment|
|Learning Diverse Rankings with Multi-Armed Bandits (2008)||2008/07/09 01:07||1 Comment|
|Deep Learning via Semi-Supervised Embedding (2008)||2008/07/06 09:56||1 Comment|
A full discussion list of all papers is also available.