ABSTRACT
Primary Progressive Aphasia (PPA) refers to a group of clinical syndromes caused by a neurodegenerative disease and is categorized into 3 types: APP-S, APP-NF/A, and APP-L. The Grammar Reception Test, in the Brazilian version (TROG2-Br), which focuses on the detection of problems in the Portuguese-Brazilian language, was used in individuals with PPA. The objective of this work is to show the feasibility of using Machine Learning algorithms to classify APP types, based on the individual's performance in the TROG-Br test. For this, the algorithms of Decision Tree, Naive Bayes, kNN, and SVM are applied and evaluated. The results are promising and show that it is possible to classify the types of APP using the individual's performance in the TROG-Br test as predictive attributes.