2018 4th International Conference on Science and Technology (ICST) · 2018 · 47 citations · 9 references
ComputingComputer ScienceArtificial Intelligence In EducationData MiningMachine LearningClassificationGenetic AlgorithmsResampling MethodsResampling TechniquesClass ImbalanceClass Imbalance HandlingClass Imbalance LearningData ResamplingImbalanceImbalanced ClassificationImbalanced LearningSynthetic Minority Oversampling
An imbalanced dataset is a condition that has a minority class which is a class has far fewer instance distributions than other classes. The imbalanced condition can affect the performance of standard classifier algorithms that lead to the biased of the results classification or tend to become a majority class. The SMOTE method overcomes the imbalanced masses by creating synthetic instances of minority classes. However, the implementation of SMOTE resulted in overgeneralization because generated instances have the same amount regardless of the distribution of instances. As a result, the boundaries between classes are unclear. The SMOTE-Simple Genetic Algorithm (SMOTE-SGA) method is used to determine the sampling rate of each instance in order to obtain unequal amounts of synthetic instances. The tests were performed using some imbalanced datasets by comparing the classification results measured using G-means and F-Measure. The results of the application of genetic algorithm at SMOTE can improve the classification result by obtaining better G-means and F-measure value.
9
SMOTE: Synthetic Minority Over-sampling Technique
Nitesh V. Chawla, Kevin W. Bowyer, Lawrence Hall et al. · Journal of Artificial Intelligence Research · 2002 · 31.2K citations · Full text
ADASYN: Adaptive synthetic sampling approach for imbalanced learning
Haibo He, Yang Bai, Edwardo A. Garcia et al. · 2008 · 4.4K citations
Cost-sensitive learning methods for imbalanced data
Nguyen Thai-Nghe, Zeno Gantner, Lars Schmidt-Thieme · 2010 · 350 citations