Evaluation of Digitalized Handwriting for Dysgraphia Detection Using Random Forest Classification Method

The paper deals with the issue of impaired hand-writing, particularly dysgraphia, and the recognition and the processing of attributes extraction. Several machine learning methods, such as random forest, support vector machine and adaptive boosting were used for this purpose. There has been 52 extracted handwriting attributes (e.g., velocity, acceleration, jerk, duration, pen lifts, etc.) from 78 handwriting samples. Only subjects aged 10–13 (inclusive 10 and 13) were included. Principal component analysis was then used in order to visualize attributes from handwriting, in two-dimensional space.