Saving newborns life through machine learning.

Saving newborns life through machine learning.
Intro: The cry of infants provides a rich source of information about the health of a newborn baby. We tried to detect asphyxia from the baby's cry. The traditional methods which are practiced are to physically check for the symptoms such as pale limbs and cold body. After, the symptoms occur doctors have very little time available for the treatment. Often, the physical methods don’t work and the child either goes into paralysis or dies.
So, I built a project which can be accessed by hospitals having a low-resource setup. The solution can be provided as an app.

I applied and extracted features from the baby’s cry using MFCC(Mel frequency cepstral coefficients) feature extraction technique.
Then, applied Support Vector Machine for the audio classification into whether the child has asphyxia. 
The accuracy was around 80%. I tried to apply machine learning but we have very limited data.