Speech Recognition for Early Detecting Alzheimer’s Disease by Using Machine Learning Algorithms

Alzheimer’s disease is a neuronal disease defined by the gradual onset of cognitive, emotional, and linguistic deficits. These assaults are severe enough to disrupt patients’ everyday social and professional lives. Nowadays, the use of speech recognition is appealing since it is non-invasive, inexpensive, and has contributed to improving accuracy. Also, it’s one of the major and innovative topics of investigation by researchers. Many other aspects may affect the accuracy of detecting Alzheimer’s including feature extraction, the number of attributes utilized for feature selection, and the classifiers used. In this paper we proposed model, which involves feature extraction and imperative attribute selection step, also classification using the machine learning algorithm support vector machine classifier. We added a new linguistic feature (silence rate) that would have the effect to increase the accuracy rate of our model. According to the current data, our proposed model can be strongly recommended for the early detection of Alzheimer’s patients from healthy people with 88% accuracy.