High density-focused uncertainty sampling for active learning over evolving stream data
Explore this paper's citation graph
Summary
This work proposes a new active learning method for evolving data streams based on a combination of density and prediction uncertainty (DBALSTREAM), which allows focusing labelling efforts in the instance space where more data is concentrated; hence the benefits of learning a more accurate classifier are expected to be higher.
- Type
- preprint
- Published
- 2014-08-24
- Cited by
- 29
- References
- 30
- OpenAlex
- https://openalex.org/W53100034
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:18192352
Keywords
Computer science, Data stream mining, Classifier (UML), Machine learning, Data stream
References
- Active Mining of Data Streams
- Using Labeled and Unlabeled Data to Learn Drifting Concepts
- Big data: The next frontier for innovation, competition, and productivity
- Extracting Hidden Context
- Handling Concept Drift in a Text Data Stream Constrained by High Labelling Cost
- Active Learning With Drifting Streaming Data
- A survey on instance selection for active learning
- Active Learning from Data Streams
- Some label efficient learning results
- A sequential algorithm for training text classifiers
- Learning model trees from evolving data streams
- Worst-Case Analysis of Selective Sampling for Linear Classification
- Relevant data expansion for learning concept drift from sparsely labeled data
- Online Active Learning Methods for Fast Label-Efficient Spam Filtering
- The Role of Hubness in Clustering High-Dimensional Data
- MOA: Massive Online Analysis
- Facing the reality of data stream classification: coping with scarcity of labeled data
- Improving Generalization with Active Learning
- Cost-sensitive online active learning with application to malicious URL detection
- An active learning system for mining time-changing data streams
Cited by
- Leading Tree in DPCLUS and Its Impact on Building Hierarchies
- Quick online spam classification method based on active and incremental learning
- Active learning over evolving data streams using paired ensemble framework
- Event detection with vector similarity based on fourier transformation
- ART: An Availability-Aware Active Learning Framework for Data Streams
- SOM-based partial labeling of imbalanced data stream
- A Novel Sampling Strategy for Active Learning over Evolving Stream Data
- Recent advances in scaling‐down sampling methods in machine learning
- Online Active Learning with Drifted Data Streams Using Paired Ensemble Framework
- Batch-based active learning: Application to social media data for crisis management
- Balanced Active Learning Method for Image Classification
- Online Active Learning Paired Ensemble for Concept Drift and Class Imbalance
- Interpreting Active Learning Methods Through Information Losses
- Fast Rotation Kernel Density Estimation over Data Streams
- Online Active Learning for Drifting Data Streams
- Stream-based active learning for sliding windows under the influence of verification latency
- Paired k-NN learners with dynamically adjusted number of neighbors for classification of drifting data streams
- Clustering-based Active Learning Classification towards Data Stream
- Ensemble Active Learning by Contextual Bandits for AI Incubation in Manufacturing
- Active learning for data streams: a survey
Related papers
- A Sliding Window Method for Finding Recently Frequent Itemsets over Online Data Streams
- Mining the Frequent Patterns in an Arbitrary Sliding Window over Online Data Streams
- Mining the Frequent Patterns in an Arbitrary Sliding Window over Online Data Streams
- Online data stream Mining of Recent Frequent Itemsets based on Sliding Window model
- A Variable Sliding Window Algorithm Based on Concept Drift for Frequent Pattern Mining Over Data Streams*
- Processing Compressed Sliding Window Continuous Queries over Data Streams