Incremental Rebalancing Learning on Evolving Data Streams

Bernardo, Alessio; Valle, Emanuele Della; Bifet, Albert

doi:10.1109/icdmw51313.2020.00121

Cited by 14 publications

(13 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Instead, the methods we propose in this paper (i.e. SML+), [18], Re-balanceStream [19] are considered active approaches. Then, there are other models like OOB and UOB [20], WEOB1 and WEOB2 [21] that, similarly to C-SMOTE, propose only an online rebalance strategy, leaving to the pipelined algorithm the concept drift management (approach type and the concept drift approach).…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Predict COVID-19 Spreading With C-SMOTE

Bernardo

Valle

2021

Bus. Inf. Sys.

Self Cite

View full text Add to dashboard Cite

Data continuously gathered monitoring the spreading of the COVID-19 pandemic form an unbounded flow of data. Accurately forecasting if the infections will increase or decrease has a high impact, but it is challenging because the pandemic spreads and contracts periodically. Technically, the flow of data is said to be imbalanced and subject to concept drifts because signs of decrements are the minority class during the spreading periods, while they become the majority class in the contraction periods and the other way round. In this paper, we propose a case study applying the Continuous Synthetic Minority Oversampling Technique (C-SMOTE), a novel meta-strategy to pipeline with Streaming Machine Learning (SML) classification algorithms, to forecast the COVID-19 pandemic trend. Benchmarking SML pipelinesthat use C-SMOTE against state-of-the-art methods on a COVID-19 dataset, we bring statistical evidence that models learned using C-SMOTE are better.

show abstract

Section: Related Workmentioning

confidence: 99%

“…We pipelined these algorithms with the C-SMOTE meta-strategy and compared them against the stand-alone versions. Instead, as SML+ models, we tested the ARF RE [18], RB [19], OOB and UOB [20] techniques. Unfortunately, the implementations of the other algorithms cited in the Related Work section were unavailable or did not work.…”

Section: Algorithmsmentioning

confidence: 99%

Predict COVID-19 Spreading With C-SMOTE

Bernardo

Valle

2021

Bus. Inf. Sys.

Self Cite

View full text Add to dashboard Cite

show abstract

“…RebalanceStream (RB) [6] uses ADWIN [8] to detect concept drift in the stream by training four models m 1 , m 2 , m 3 , and m 4 in parallel: m 1 is trained with the original samples in input; m 2 uses the samples collected and rebalanced using SMOTE from the beginning to a change (when the last concept drift occurred); m 3 uses the samples collected from a warning (when the most recent concept drift started to occur) to a change; and m 4 uses the same data of m 3 but rebalanced using SMOTE. After a change, the best model among them is chosen and it is used to continue the execution according to k-statistics.…”

Section: Related Workmentioning

confidence: 99%

“…We pipeline these algorithms with the new C-SMOTE meta-strategy and compare against the stand-alone versions. We also compare the ARF RE [5] technique to ARF pipelined with C-SMOTE and the RB [6] strategy to SWT pipelined with C-SMOTE. We show the OOB and UOB [7] results compared to the best technique pipelined with C-SMOTE, too.…”

Section: A Experimental Settingsmentioning

confidence: 99%

“…• C-SMOTE, a meta-strategy inspired by the well known SMOTE technique (applicable only to batches) that can be pipelined with any data stream classifier; • an analysis of the performances of C-SMOTE when pipelined with different data stream classifiers on ten groups of dynamic and imbalanced data streams (eight real and two synthetic) showing at least an improvement of the minority class performances; and • a comparison between C-SMOTE and four other strategies able to deal with class imbalance in the streaming scenario, namely Adaptive Random Forest with Resampling (ARF RE ) [5], RebalanceStream (RB) [6], undersampling-based Online Bagging (UOB) and oversampling-based Online Bagging (OOB) [7] showing that there is at least one algorithm pipelined with C-SMOTE that outperforms the state of the art one. The remainder of this paper is organized as follows.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

C-SMOTE: Continuous Synthetic Minority Oversampling for Evolving Data Streams

Bernardo

Gomes

Montiel

et al. 2020

2020 IEEE International Conference on Big Data (Big Data)

Self Cite

View full text Add to dashboard Cite

Streaming Machine Learning (SML) studies singlepass learning algorithms that update their models one data item at a time given an unbounded and often non-stationary flow of data (a.k.a., in presence of concept drift). Online class imbalance learning is a branch of SML that combines the challenges of both class imbalance and concept drift. In this paper, we investigate the binary classification problem of rebalancing an imbalanced stream of data in the presence of concept drift, accessing one sample at a time. We propose Continuous Synthetic Minority Oversampling Technique (C-SMOTE), a novel rebalancing meta-strategy to pipeline with SML classification algorithms. C-SMOTE is inspired by the popular SMOTE algorithm but operates continuously. We benchmark C-SMOTE pipelines on ten different groups of data streams. We bring empirical evidence that models learnt with C-SMOTE pipelines outperform models trained on imbalanced data stream without losing the ability to deal with concept drifts. Moreover, we show that they outperform other stream balancing techniques from the literature.

show abstract