Two-stage RFID approach for localizing objects in smart homes based on gradient boosted decision trees with under-and over-sampling
S Abudalfa, K Bouchard - Journal of Reliable Intelligent Environments, 2024 - Springer
Developing automated systems with a reasonable cost for long-term care for elders is a
promising research direction. Such smart systems are based on realizing activities of daily …
promising research direction. Such smart systems are based on realizing activities of daily …
CCO: A Cluster Core-Based Oversampling Technique for Improved Class-Imbalanced Learning
P Mondal, F Ansari, S Das - IEEE Transactions on Emerging …, 2024 - ieeexplore.ieee.org
Supervised classification problems from the real world typically face a challenge
characterized by the scarcity of samples in one or more target classes compared to the rest …
characterized by the scarcity of samples in one or more target classes compared to the rest …
Addressing Class Imbalance of Health Data: A Systematic Literature Review on Modified Synthetic Minority Oversampling Technique (SMOTE) Strategies
Abstract The Synthetic Minority Oversampling Technique (SMOTE) method is the baseline
for solving unbalanced data problems. The working concept of the SMOTE method is to …
for solving unbalanced data problems. The working concept of the SMOTE method is to …