Dissertation Information for Mollie PoyntonNAME:
DEGREE:
DISCIPLINE:
SCHOOL: ADVISORS: COMMITTEE MEMBERS: MPACT Status: Incomplete - Not_Inspected Title: Classification of smoking cessation status with machine learning methods Abstract: This study examined the ability of a backpropagation neural network (BPNN) classifier to distinguish between current and former smokers in the 2000 National Health Interview Survey (NHIS) sample adult file. Pre-processing and automated attribute subset selection aided in reduction of the search space, or number of possible solutions. A multilayer artificial neural network was constructed and trained to predict class membership, current or former smoker, using inputs from the attribute subset. Classifier performance exceeded that of random chance, with Az asymmetric 95% confidence intervals (0.7532, 0.7790). However, separation of current and former smokers was imperfect, as illustrated by the ROC (receiver operating characteristic) curve. Additionally, performance did not exceed that of a comparison classifier created using logistic regression. These findings indicate that imprecise, but high quality health survey data provided sufficient information for prediction of smoking cessation status. Membership of specific survey items in the attribute subset was largely supported by the literature. However, the inability to interpret relationships among attribute values in the BPNN model hampers comparison with published literature. The study establishes the ability of a machine learning method, backpropagation neural networks, to classify a health-related behavior, smoking cessation. |
MPACT Scores for Mollie PoyntonA = 0 Advisors and Advisees Graph |