THANK YOU FOR SUBSCRIBING
The use of Asynchronous Data Augmentation creates the possibility of quick and easy generation of artificial structures for automatic validation of educational operations.
FREMONT, CA: Evaluating student progress at each learning stage in an individualized process is extremely complex and arduous. The only remedy is to automate evaluation using Deep Learning methods. The challenge is the relatively small amount of data, in the form of available evaluations, required to train the neural network. The specificity of each subject taught needs the preparation of a separate neural network. Here is a new data augmentation method, Asynchronous Data Augmentation through Pre-Categorization (ADAPC), which resolves this problem. It is possible to train an effective deep neural network with the proposed means, even for a small amount of data.
Deep Learning (DL) in teaching began to spread recently. In recent years, a massive increase in the use of neural networks in teaching has been seen in student evaluation automation. Two sectors that automation applies to can be found. The first is automated scoring, and the second to automatic grading, automatically classifying student responses based on a previous correct answer. Particularly interesting are attempts to use DL potentials in the field of text analysis. Methods based on the application of recurrent neural networks, including bidirectional LTSM networks, dominate here.
The priority of modern education is to implement the means and pace of knowledge and skills transfer to each student's predispositions. Such a strategy needs both the division of the where learning process into small multi-variant stages and the evaluation of mastery of knowledge at the end of each stage. It is possible to revamp the course of the teaching process for each student by using evaluation carried out in stages. Such multi-variability of the option of the further educational path is essential– choosing the type of next stage from among many options available, based on the result of the previous stage's assessment.