手法を比較
選択した手法を並べて確認できます。異なる行はハイライト表示されます。
| アクティブラーニング・ガウス混合モデル× | アクティブラーニング・ガウシアンプロセス× | |
|---|---|---|
| 分野 | 機械学習 | 機械学習 |
| 系統 | Machine learning | Machine learning |
| 提唱年≠ | 2000s (combination) | 1992 |
| 提唱者≠ | Settles, B. (active learning framework); Dempster, Laird & Rubin (GMM via EM, 1977) | MacKay, D. J. C. |
| 種類≠ | Active learning for probabilistic clustering / density estimation | Bayesian active learning |
| 原典≠ | Zhu, X., Ghahramani, Z., & Lafferty, J. (2003). Semi-supervised learning using Gaussian fields and harmonic functions. Proceedings of the 20th International Conference on Machine Learning (ICML), 912–919. link ↗ | MacKay, D. J. C. (1992). Information-based objective functions for active data selection. Neural Computation, 4(4), 590–604. DOI ↗ |
| 別名 | AL-GMM, active GMM, query-by-committee GMM, active density estimation | GP active learning, Gaussian process active learning, GP-AL, Bayesian active learning with GP |
| 関連 | 4 | 4 |
| 概要≠ | Active Learning Gaussian Mixture Model combines an iterative query strategy with a Gaussian Mixture Model learner. The algorithm selects the most informative unlabeled points — typically those with highest predictive uncertainty — presents them to an oracle for labeling, and refits the GMM using EM on the growing labeled set. The result is a density model that matches full-data quality while requiring far fewer labeled examples. | Active Learning Gaussian Process (GP-AL) combines a Gaussian process probabilistic model with an active learning query strategy, using the GP's posterior uncertainty to select the most informative unlabeled examples for labeling. This iterative approach minimizes labeling effort while maximizing predictive accuracy, making it ideal when labeled data is scarce or expensive to obtain. |
| ScholarGateデータセット ↗ |
|
|