Passive Concept Drift Handling via Momentum Based Robust Soft Learning Vector Quantization

Heusinger, Moritz; Raab, Christoph; Schleif, Frank-Michael

doi:10.1007/978-3-030-19642-4_20

Cited by 10 publications

(12 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…6is not computed at the update step and, therefore, the RSLVQ is feasible for potentially infinite streams. The prototypes in each update step will be optimized with a momentumbased gradient technique designed for RSLVQ [21] given with…”

Section: Robust Soft Learning Vector Quantizationmentioning

confidence: 99%

See 1 more Smart Citation

Reactive Soft Prototype Computing for Concept Drift Streams

2020

Self Cite

View full text Add to dashboard Cite

The amount of real-time communication between agents in an information system has increased rapidly since the beginning of the decade. This is because the use of these systems, e.g. social media, has become commonplace in today's society. This requires analytical algorithms to learn and predict this stream of information in real-time. The nature of these systems is non-static and can be explained, among other things, by the fast pace of trends. This creates an environment in which algorithms must recognize changes and adapt. Recent work shows vital research in the field, but mainly lack stable performance during model adaptation. In this work, a concept drift detection strategy followed by a prototype-based adaptation strategy is proposed. Validated through experimental results on a variety of typical non-static data, our solution provides stable and quick adjustments in times of change.

show abstract

Section: Robust Soft Learning Vector Quantizationmentioning

confidence: 99%

“…The σ is the width of the Gaussian kernel and together with the number of prototypes, are the only tuneable parameters in the RSLVQ. For a more comprehensive derivation of RSLVQ with momentum SGD see [3,21].…”

Section: Robust Soft Learning Vector Quantizationmentioning

confidence: 99%

Reactive Soft Prototype Computing for Concept Drift Streams

2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…Also, LVQ algorithms first introduced by [19] have received attention as potential stream classification algorithms [30]. The Robust Soft Learning Vector Quantization (RSLVQ) [29] is a promising probabilistic classifier, which assumes class distributions as Gaussian mixture models learned via Stochastic Gradient Ascent (SGA) and has only been evaluated as a stream classifier in a previous version of this article [16] so far. Also, Generalized Learning Vector Quantization has not been considered as a stream classifier yet.…”

Section: Related Workmentioning

confidence: 99%

“…Hence, we do not optimize hyperparameters in our experiments. Note that in [16] hyperparameters of all classifiers were tuned, which lead to different results.…”

Section: Setupmentioning

confidence: 99%

“…In [30], it was also shown that the usage of weight decay has the potential to improve the generalization error of the LVQ in non-stationary environments. Other LVQ variants have not been considered as stream classifiers yet, except from our adaptive Robust Soft Learning Vector Quantization (RSLVQ) in [16]. Therefore, we propose adaptive versions of RSLVQ and Generalized Learning Vector Quantization (GLVQ) which maximize the objective function with momentum-based gradient descent/ascent in Sect.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Passive concept drift handling via variations of learning vector quantization

Heusinger

Raab

Schleif

2020

Neural Comput & Applic

Self Cite

View full text Add to dashboard Cite

Concept drift is a change of the underlying data distribution which occurs especially with streaming data. Besides other challenges in the field of streaming data classification, concept drift has to be addressed to obtain reliable predictions. Robust Soft Learning Vector Quantization as well as Generalized Learning Vector Quantization has already shown good performance in traditional settings and is modified in this work to handle streaming data. Further, momentum-based stochastic gradient descent techniques are applied to tackle concept drift passively due to increased learning capabilities. The proposed work is tested against common benchmark algorithms and streaming data in the field and achieved promising results.

show abstract

Random Projection in the Presence of Concept Drift in Supervised Environments

Heusinger

Schleif

2020

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Passive Concept Drift Handling via Momentum Based Robust Soft Learning Vector Quantization

Cited by 10 publications

References 16 publications

Reactive Soft Prototype Computing for Concept Drift Streams

Reactive Soft Prototype Computing for Concept Drift Streams

Passive concept drift handling via variations of learning vector quantization

Random Projection in the Presence of Concept Drift in Supervised Environments

Contact Info

Product

Resources

About