An Empirical study and assessment of minority oversampling with Dynamic Ensemble Selection on COVID-19 utilizing Blood Sample

The COVID-19 virus disease outbreak that erupted in China at the end of 2019 has had a tremendous and disastrous impact on the rest of the world. It has struck the globe to its core, and the destruction has substantially increased the diagnostic burden. In the pandemic zone, clinicians will be able to cut down on their workload and get the right diagnosis of the new disease great to the use of machine learning. A blood test has emerged as a critical tool for identifying false-positive or false-negative real-time rRT-PCR diagnostics. Notably, this is mostly because it is such a cost-effective and convenient method of detecting probable COVID-19 patients. Among the numerous hard consequences associated with COVID-19 illness has been established as one of the most prevalent among COVID-19 patients. The impetus for this research is the scarcity of post-COVID-19 dataset. Following pre-processing to manage address missing values, oversampling with SMOTE ENN is used to generate several instances and model training is carried out on these data sets. However, it has been demonstrated that normatively dynamic ensemble selection outperforms static selection and dynamic selection. The DI+SMOTEENN+DESKNU exceed existing benchmark Classification algorithms and obtain the best accuracy of 99.6%, according the results. © 2022 IEEE.

A novel combined dynamic ensemble selection model for imbalanced data to detect COVID-19 from complete blood count.

BACKGROUND: As blood testing is radiation-free, low-cost and simple to operate, some researchers use machine learning to detect COVID-19 from blood test data. However, few studies take into consideration the imbalanced data distribution, which can impair the performance of a classifier. METHOD: A novel combined dynamic ensemble selection (DES) method is proposed for imbalanced data to detect COVID-19 from complete blood count. This method combines data preprocessing and improved DES. Firstly, we use the hybrid synthetic minority over-sampling technique and edited nearest neighbor (SMOTE-ENN) to balance data and remove noise. Secondly, in order to improve the performance of DES, a novel hybrid multiple clustering and bagging classifier generation (HMCBCG) method is proposed to reinforce the diversity and local regional competence of candidate classifiers. RESULTS: The experimental results based on three popular DES methods show that the performance of HMCBCG is better than only use bagging. HMCBCG+KNE obtains the best performance for COVID-19 screening with 99.81% accuracy, 99.86% F1, 99.78% G-mean and 99.81% AUC. CONCLUSION: Compared to other advanced methods, our combined DES model can improve accuracy, G-mean, F1 and AUC of COVID-19 screening.