Evaluate student achievement by classifying brain structure and its functionality with novel hybrid method
Künye
Ataş, P. K. (2023). Evaluate student achievement by classifying brain structure and its functionality with novel hybrid method. Neural Computing and Applications, 1-12.Özet
In a labor market that demands a workforce well-trained in science, technology, engineering, and mathematics (STEM) subjects, it is required of children to successfully develop their mathematical skills in order to become highly productive adults. Recent developments in computer vision, artificial intelligence, machine learning, and medical imaging techniques give us new opportunities for building intelligent support tools to help us learn more about the neural underpinnings of how children learn math and how that knowledge relates to individual differences in skill. This study examines the brain activities of students during problem-solving by checking brain structure and its functionality. By using powerful techniques in the light of machine learning and image processing, the relationship between success and the background of a child was researched. The aim is to make a solid prediction of the possible future success of the children by observing their brain activities. The children we investigated were asked different questions to get information about their intelligence. In our study, we have tried to find how those questions and answers may affect the future success of a child. For this purpose, a novel hybrid classification model that utilizes cluster analysis, Random Forest, Logistic Regression, and ensemble learning is intended for classification tasks. Our study includes two main stages. Firstly, the image processing techniques were applied to create unique features of brain images. Then, machine learning tecnniques were used to select a set of features, and for getting prediction results our hybrid classification model was applied. In the end, we obtained useful results indicating that there is a complicated connection between the success rate and the history of a child. This novel approach to classification, which combines multiple methods by using a hybrid model, has the potential to be implemented in computational tools for strategic decision support systems.