Can you Understand why I am Crying? A Decision-making System for Classifying Infants' Cry Languages Based on deepSVM Model

There scientific and therapeutic advances in perinatology neonatology have improved the survival prospects of preterm extremely low birth weight infants. Infants' cries are a valuable noninvasive tool for monitoring their neurologic health, especially if they premature. An automatic acoustic analysis data mining employed this study to determine discriminative features full-term infant cries. The use machine learning recognizing sounds newborn's cry language has received less attention than previous methods analyzing sounds. Moreover, extract appropriate from cries, adequate knowledge signal descriptors required. Accordingly, analyze language, we propose an approach that uses fractal descriptor discriminant spectrograms windowed signals, followed by iterative neighborhood component (iNCA) select features. Additionally, deep support vector (deepSVM) is used classify infants' crying types meanings. proposed method verified using newborn sound dataset. According classification five perception based on various characteristics, 98.34% all perceptions been recognized. Although there many classes examined, feature extraction our optimal much higher diagnostic accuracy compared similar baby language. can overcome problems associated with babies' understanding like uncertainty unusual errors classification.