Machine Learning Model to Predict the ADHD Patient from the Text Content
JEONG, Seyoon
DOI: http://doi.org/10.34614/2022IYRC22
Category: STEM
Abstract – Regarding all the possible innovations that relate to diverse methods including classification and analysis of data, a written text may perform as a valuable resource to inspect one’s symptoms or standing traits. In connection with diagnosing a mental disease or indications implying a specific condition, texts, in any form, can be utilized as a new method of diagnosing and treating patients. Specifically, as people write genuine opinions and comments on the Internet, written data is one efficient source for classification programs. Considering the suggestions of designing a computer model to analyze and determine the diagnosis of mental illness: ADHD, this creation will benefit those who may not have been in favor of taking the economic and social risks of diagnosing the disease in hospitals. Utilizing different types of classification programs and text-analysis machine techniques, the computer model would function as a tool to easily determine possibilities of ADHD with authentic data sets of the patient from Internet records. Continuing the project after designing and testing the computer models made, models were plotted into a graph to compare their accuracies for correctly analyzing the data text. As a result, LSTM rated highest with an accuracy of 0.75, followed by RNN with a score of 0.65.