Multi-class Classification of Imbalanced Intelligent Data using Deep Neural Network

In recent years, studies in the field of deep learning have made significant progress. These focusedmore on datasets with balanced classification, and less research has been done imbalanced datasets, whichare great importance real world present challenges for classification. This articlestudies problem classifying data, introduces dynamic sampling neural networks,investigates multiclass problem, proposes a method learning.In our proposed method, all samples are fed to current network each training iteration,and accuracy, precision, mean error estimated. The methoddynamically selects informative data network. Comprehensive experiments wereconducted evaluate understand its strengths weaknesses. results 13 multiclassdatasets show that outperforms other methods, such as initial techniques,active learning, cost-sensitive reinforcement learning.