Incremental learning from noisy data

By Jeffrey C. Schlimmer, Richard Granger. 2004

Subfields: Optimization & Training, Meta & Continual Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Optimization & Training

In Meta & Continual Learning

Citations in the corpus

Missing Data Imputation for Supervised Learning (2016, methodology): “The CVRs dataset donor claims to achieve a 5-10% error rate using an incremental decision tree algorithm called STAGGER, although it is unknown to the authors what train-test split is used or how missing values are handled (Schlimmer, 1987; Schlimmer and Granger Jr, 1986).”
Recent Advances of Continual Learning in Computer Vision: An Overview (2021, neutral): “Secondly, when learning from new data of the same category in sequential order, this can also lead to the problem of concept drift [3–5], as the new data may change the data distribution of this category in unforeseen ways [6].”
IIRC: Incremental Implicitly-Refined Classification (2020, background): “Lifelong Learning is a broad, multi-disciplinary, and expansive research domain with several synonyms: Incremental Learning [42], Continual Learning [48], and Never Ending Learning [35].”
Continual Lifelong Learning in Natural Language Processing: A Survey (2020, background): “Models become less accurate over time, for instance, due to the changing distribution of data – the phenomenon known as concept drift (Schlimmer and Granger, 1986; Widmer and Kubat, 1993).”
Learn$^{++}$ .NC: Combining Ensemble of Classifiers With Dynamically Weighted Consult-and-Vote for Efficient Incremental Learning of New Classes (2009, methodology): “tion of classifier weights [27], [28], or architecture update by growing/pruning of decision trees, nodes or clusters [29]–[31], or both as in STAGGER [32] and hybrid decision trees [5].”
The Impact of Diversity on Online Ensemble Learning in the Presence of Concept Drift (2010, unknown): “concept drift benchmarks (STAGGER Boolean concepts [42]”
MOA: Massive Online Analysis, a Framework for Stream Classification and Clustering (2010, unknown): “STAGGER Concepts Generator They were introduced by Schlimmer and Granger in [39].”
Continuous Domain Adaptation with Variational Domain-Agnostic Feature Replay (2020, background): “…machine learning is to learn in non-stationary environments, in which the underlying data distribution (i.e., the joint distribution of the input data and labels P (X,Y )) changes over time, also referred to as concept drift in previous literature (Schlimmer & Granger, 1986; Widmer & Kubat, 1996).”
Learning under Concept Drift: an Overview (2010, background): “Schlimmer and Granger [136] in 1986 formulated the problem of incremental learning from noisy data and presented an adaptive learning algorithm STAGGER.”
Continual Learning with Gated Incremental Memories for sequential data processing (2020, methodology): “Similar to the approach proposed in this paper, ensemble methods associate an expert to each (or to a group of) concepts and combine their predictions into the final answer [12, 13].”
Recognition From Web Data: A Progressive Filtering Approach (2018, unknown): “The strategy of progressive learning has been used in data mining [20], pattern recognition [21], computer vision [22]–[24], etc .”
Applying lazy learning algorithms to tackle concept drift in spam filtering (2007, background): “Such changes are usually referred to as concept drift (Schlimmer & Granger, 1986) and can be caused by a changed context.”
A framework for generating data to simulate changing environments (2007, methodology): “The classical example is the STAGGER concept first proposed by Schlimmer and Granger [28].”