Date Presented

Spring 5-15-2022

Document Type

Thesis

Degree Name

Bachelor of Science

Department

Computer Science

First Advisor

Dr. Zakaria Kurdi

Second Advisor

Dr. Randy Ribler

Third Advisor

Dr. Laura Kicklighter

Abstract

The goal of this work is to build a classifier that can identify whether a patient is suffering from Alzheimer’s Disease of the Dementia Type (AD). A corpus of 2751 texts was used from the DementiaBank database, where each conversation is transcribed and marked using the CHAT format. Each text was analyzed by frequency of disfluencies, use of aphasic language, and lexical features. All parsed data was used to train a Random Forest, Naïve Bayes, and Support Vector Machine algorithm. These classification algorithms will be tested on the combination of all features, as well as each set of features individually.

Share

COinS