LAUSR

The active learning method involves searching for the most informative unmarked samples by query function, submitting them to the expert function for marking, then using the samples to train the classification model in order to improve the accuracy of the model and use the newly acquired knowledge to inquire into the next round, with the aim of getting the highest accuracy of classification using minimal training samples. This paper details the various principles of active learning and develops a method that combines active learning with transfer learning. Experimental results prove that the active learning method can cut back on samples redundancy and promote the accuracy of classifier convergence quickly in small samples. Combining active learning and transfer learning, while taking advantage of knowledge in related areas, could further improve the generalization ability of classification models.