Active Learning by Querying Informative and Representative Examples

Huang, Sheng-Jun; Jin, Rong; Zhou, Zhi‐Hua

doi:10.1109/tpami.2014.2307881

Cited by 411 publications

(382 citation statements)

References 30 publications

Supporting

Mentioning

371

Contrasting

Unclassified

Order By: Relevance

“…The global performance displayed on figures results from averaging the accuracy of every run. We compare OLRM and OGRM with the following methods: (1) Random Sampling: each instance to be labeled is randomly drawn from the pool, (2) Uncertainty Sampling (US) [6] samples instance closest to the boundary, (3) QUIRE [3] min-max strategy for minimization of the global risk, (4) VOI [5] minimization of the expected global risk, and (5) GPAL (Uncertainty) [4] minimization of the expected local risk.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Optimism in Active Learning with Gaussian Processes

Collet

Pietquin

2015

Neural Information Processing

View full text Add to dashboard Cite

Abstract. In the context of Active Learning for classification, the classification error depends on the joint distribution of samples and their labels which is initially unknown. The minimization of this error requires estimating this distribution. Online estimation of this distribution involves a trade-off between exploration and exploitation. This is a common problem in machine learning for which multi-armed bandit theory, building upon Optimism in the Face of Uncertainty, has been proven very efficient these last years. We introduce two novel algorithms that use Optimism in the Face of Uncertainty along with Gaussian Processes for the Active Learning problem. The evaluation lead on real world datasets shows that these new algorithms compare positively to state-of-the-art methods.

show abstract

Section: Methodsmentioning

confidence: 99%

“…In [4] and [5], the binary loss is expected over the possible values of the true conditional density of the class label. In [3], a min-max approach is used to ensure a minimum decrease of the risk.…”

Section: Introductionmentioning

confidence: 99%

Optimism in Active Learning with Gaussian Processes

Collet

Pietquin

2015

Neural Information Processing

View full text Add to dashboard Cite

show abstract

“…The other one directly seeks such a task by estimating the distribution of possible states for a given task. Based on this idea, we propose an active sampling strategy incorporating the active learning [14] idea. For an unseen task, we employ an SVM model to estimate the confidence that the weak policy will be successful, a Gaussian Mixture Model to estimate the distance to the explored area in task space, and the likelihood that unseen states will be visited.…”

Section: Active Samplingmentioning

confidence: 99%

Self-Practice Imitation Learning from Weak Policy

Zhou

2013

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Abstract. Imitation learning is an effective strategy to reinforcement learning, which avoids the delayed reward problem by learning from mentor-demonstrated trajectories. A limitation for imitation learning is that collecting sufficient qualified demonstrations is quite expensive. In this work, we study how an agent can automatically improve its performance from a weak policy, by automatically acquiring more demonstrations for learning. We propose the LEWE framework to sample tasks for the weak policy to execute, and then learn from the successful trajectories to achieve an improvement. As the sampling strategy is the key to the efficiency of LEWE, we further propose to incorporate active learning for the sampling strategy for LEWE. Experiments in a spatial positioning task show that LEWE with active learning can effectively and efficiently improve the weak policy and achieves a better performance than the comparing sampling approaches.

show abstract

“…Note that how to form a query plays a key role in an active learning algorithm. In terms of query formation, there are two scenarios of active learning in the literature: query synthesis [5,6,7] and sampling, which can be further divided into stream-based sampling [8,9] and pool-based sampling [10,3,11,12] .…”

Section: Introductionmentioning

confidence: 99%

Active learning via query synthesis and nearest neighbour search

Wang

Yuan

et al. 2015

Neurocomputing

View full text Add to dashboard Cite

Active learning has received great interests from researchers due to its ability to reduce the amount of supervision required for effective learning. As the core component of active learning algorithms, query synthesis and pool-based sampling are two main scenarios of querying considered in the literature. Query synthesis features low querying time, but only has limited applications as the synthesized query might be unrecognizable to human oracle. As a result, most efforts have focused on pool-based sampling in recent years, although it is much more time-consuming. In this paper, we propose new strategies for a novel querying framework that combines query synthesis and pool-based sampling. It overcomes the limitation of query synthesis, and has the advantage of fast querying. The basic idea is to synthesize an instance close to the decision boundary using labeled data, and then select the real instance closest to the synthesized one as a query. For this purpose, we propose a synthesis strategy, which can synthesize instances close to the decision boundary and spreading along the decision boundary. Since the synthesis only depends on the relatively small labelled set, instead of evaluating the entire unlabeled set as many other active learning algorithms do, our method has the advantage of efficiency. In order to handle more complicated data and make our framework compatible with powerful kernel-based learners, we also extend our method to kernel version. Experiments on several real-world data sets show that our method has significant advantage on time complexity and similar performance compared to pool-based uncertainty sampling methods.

show abstract

Active Learning by Querying Informative and Representative Examples

Cited by 411 publications

References 30 publications

Optimism in Active Learning with Gaussian Processes

Optimism in Active Learning with Gaussian Processes

Self-Practice Imitation Learning from Weak Policy

Active learning via query synthesis and nearest neighbour search

Contact Info

Product

Resources

About