AI added multinationalists Wolf
Researchers from the Institute of North Carolina (North Carolina State University) developed a new model of artificial intelligence (AI) that helps to improve the learning process of students and
enhancing its effectiveness.
Advanced model of artificial intelligence is characterized in that the analysis of the effectiveness of training checks the behavior of the student, that is how he copes with the stages of the educational gaming
programs, in the context of the results for each test question, not the total score, as do traditional algorithms. The studies used gaming training program
Crystal Island.
Scientists have tried to make AI based on the experience of students that have passed the test, was able to predict – how will answer a particular test question a student who is still undergoing training.
“The standard approach to solving this problem looks only at the total test score, considering the test as a single task,” explained research specialist University Jonathan Rowe (Jonathan Rowe). – “In
the context of our structure of multitask learning in the model 17 task, because in a test of 17 questions.”
The researchers obtained data on training obtained during testing 181 of the student. AI could see for himself the learning process of each student and how each student answered
a particular test question. Identifying General patterns in the behavior of students who correctly answered the selected question, as well as the typical behavior of students who answered it incorrectly, AI
could do with a good accuracy to determine which answer is the student who is still undergoing training.
“We assume that we have developed a model type AI can be used in several ways that can benefit students,” said Professor Michael Eden (Michael Geden). – “It
you can apply for a notice to teachers that in the process of training the student requires additional instruction and clarification. In addition, AI can help to improve the educational
is the process”.
“Psychology has long recognized that different questions have different values, – said Michael Gaiden. – Our work is based on an interdisciplinary approach that combines this aspect of psychology with the approaches to
more deep machine learning”.
Andrew Emerson (Andrew Emerson), Ph. D. from the University of North Carolina believes that the achievements made by him and his colleagues, open doors to the more sophisticated methods
modelling in educational software, allowing you to tailor the learning process to the needs of the learner.
Artificial intelligence
