The goal was to develop a statistical model that could use dysphonia measures to replicate speech experts’ assessment of future rehabilitation treatment status. I constructed a neural network classifier that utilized principal component analysis and resilient backpropagation with weight tracking. Final model performed with 94% test accuracy.
andreioprisan / parkinsons_classifier Goto Github PK
View Code? Open in Web Editor NEWThis project forked from eshiang21/parkinsons_classifier