Tree-Based Ensemble Multi-Task Learning Method for Classification and Regression
Jaak Simm, I. Magrans, Masashi Sugiyama · IEICE Transactions on Information and Systems · 2014
Multi-task learning is an important area of machine learning that tries to learn multiple tasks simultaneously to improve the accuracy of each individual task. We propose a new tree-based ensemble multi-task learning method for classification and regression (MT-ExtraTrees), based on Extremely Randomized Trees. MT-ExtraTrees is able to share data between tasks minimizing negative transfer while keeping the ability to learn non-linear solutions and to scale well to large datasets.
Cited by 83 works. doi:10.1587/transinf.e97.d.1677 Open-access PDF
Authors
References
- Random Forests
- Improved boosting algorithms using confidence-rated predictions
- Improved Boosting Algorithms Using Confidence-rated Predictions
- Extremely randomized trees
- Regularized multi--task learning
- Multi-task learning for classification with dirichlet process priors
- MULTIBOOST: a multi-purpose boosting package
- Computationally Efficient Multi-task Learning with Least-squares Probabilistic Classifiers
- Learning Multiple Tasks with Boosted Decision Trees
- A Hierarchical Approach for Multi-task Logistic Regression