Imbalanced data in one of the major problems faced by Machine learning and deep learning classifiers. The skewness in the data distribution limits the performance of classifiers. This leads to overfitting of the model and misclassification for minority classes. Researchers have been focused on new techniques to balance data by oversampling minority classes, under sampling majority classes or creating a hybrid of oversampling and under sampling. Over the years researchers have also explored algorithmic techniques to adjust weights, create bags of classes and optimally enhance the data. This paper provides a stateof-the-art review of the latest contributions to resolve the imbalance data problem. The major focus of this paper is on the hybrid techniques, ensemble methods and GAN-based data augmentation techniques