A cluster-based intelligence ensemble learning method for classification problems

Cui, Shaoze; Wang, Yanzhang; Yin, Yunqiang; Cheng, Tao; Wang, Dujuan; Zhai, Mingyu

doi:10.1016/j.ins.2021.01.061

Cited by 25 publications

(8 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ensemble methods [14,15] are based on creating several models from the training data. They have been postulated as an efficient alternative for complex problems, where the construction of different models from the data, in such a way that they complement each other, usually brings some advantages [49].…”

Section: Building Classification Ensembles Using Baggingmentioning

confidence: 99%

“…, p t . The most used approach for output combination in the specialized literature is majority voting [14,19]. This is a simple but effective procedure in which each model within the ensemble casts a vote for one of the classes, and the most voted class is chosen as the final prediction.…”

Section: Building Classification Ensembles Using Baggingmentioning

confidence: 99%

“…When dealing with complex datasets, several works have demonstrated that ensemble methods [14,15], which simultaneously use several classifiers, are an accurate way to overcome some of the difficulties in building models from the data. One of the best known approaches to create ensembles is bootstrap aggregating or, as it is more commonly known, bagging [16,17].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

On the Suitability of Bagging-Based Ensembles with Borderline Label Noise

Sáez

Romero‐Béjar

2022

Mathematics

View full text Add to dashboard Cite

Real-world classification data usually contain noise, which can affect the accuracy of the models and their complexity. In this context, an interesting approach to reduce the effects of noise is building ensembles of classifiers, which traditionally have been credited with the ability to tackle difficult problems. Among the alternatives to build ensembles with noisy data, bagging has shown some potential in the specialized literature. However, existing works in this field are limited and only focus on the study of noise based on a random mislabeling, which is unlikely to occur in real-world applications. Recent research shows that other types of noise, such as that occurring at class boundaries, are more common and challenging for classification algorithms. This paper delves into the analysis of the usage of bagging techniques in these complex problems, in which noise affects the decision boundaries among classes. In order to investigate whether bagging is able to reduce the impact of borderline noise, an experimental study is carried out considering a large number of datasets with different noise levels, and several noise models and classification algorithms. The results obtained reflect that bagging obtains a better accuracy and robustness than the individual models with this complex type of noise. The highest improvements in average accuracy are around 2–4% and are generally found at medium-high noise levels (from 15–20% onwards). The partial consideration of noisy samples when creating the subsamples from the original training set in bagging can make it so that only some parts of the decision boundaries among classes are impaired when building each model, reducing the impact of noise in the global system.

show abstract

Section: Building Classification Ensembles Using Baggingmentioning

confidence: 99%

Section: Building Classification Ensembles Using Baggingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

On the Suitability of Bagging-Based Ensembles with Borderline Label Noise

Sáez

Romero‐Béjar

2022

Mathematics

View full text Add to dashboard Cite

show abstract

“…In addition, we propose a two-layer nested heterogeneous ensemble learning-based prediction method to improve the prediction accuracy of COVID-19 mortality. If we can develop a prediction model that can accurately judge the changing trend of COVID-19 mortality in the future, it will play an auxiliary decision-making role in the formulation of policies and deployment of rescue resources to combat COVID- 19.…”

Section: Two-layer Nested Heterogeneous Ensemble Learning-based Predi...mentioning

confidence: 99%

“…Among the various machine learning methods, the ensemble learning method is the most popular in recent years, which has been successfully applied in many fields such as medical diagnosis [16], risk assessment [17], and fault diagnosis [18]. The ensemble learning method generates the final prediction result by combining the outputs of multiple base learners [19]. In this way, the fault tolerance of the entire model can be effectively improved and the generalization error can be reduced.…”

Section: Introductionmentioning

confidence: 99%