Grafik Tablet Kullanılarak Makine Öğrenmesi Yardımı ile El Yazısından Cinsiyet Tespiti.

As a routine of daily life handwriting gives clues about mood, personality traits and some diseases . In addition, many disciplines such as forensic medicine, medicine and archeology often use handwriting in their own fields. For example, forensic medicine can reach information from the handwriting, such as age range and which hand is used in some cases. In this study, a handwriting system was proposed. The proposed system subtracted a series of features from the handwriting that would characterize the handwriting, classifying them according to gender that used in machine learning techniques. With the tablet in which handwriting is recorded, the feature is extracted from the curve movements that the pen follows in the air both in case the pen is in contact with the tablet and in the transition between letters and words. These features are the speed of the pen, acceleration, jerking movements while writing, the angle of inclination, the curves in the writing, the rate of pen stay in the air, the pressure value of the pen and the height angle of the pen. Decision Trees (DT), Naive Bayes (NB), Support Vector Machines (SVM) and k-Nearest Neighbor (k-NN) approaches were used as the classifier. There are 410 samples in the data set used in the experimental studies and the performances of the experimental studies were evaluated with accuracy criterion. According to the results that have been observed, the best performance was obtained with SVM and the accuracy value was 85.1%. [ABSTRACT FROM AUTHOR]