CCE Theses and Dissertations

Date of Award

2020

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

College of Computing and Engineering

Advisor

Sumitra Mukherjee

Committee Member

Michael J. Laszlo

Committee Member

Francisco J. Mitropoulos

Abstract

Training supervised machine learning models requires labeled examples. A judicious choice of examples is helpful when there is a significant cost associated with assigning labels. This dissertation aims to improve upon a promising extant method - Batch-mode Expected Model Change Maximization (B-EMCM) method - for selecting examples to be labeled for regression problems. Specifically, it aims to develop and evaluate alternate strategies for adaptively selecting batch size in B-EMCM, named adaptive B-EMCM (AB-EMCM).

By determining the cumulative error that occurs from the estimation of the stochastic gradient descent, a stop criteria for each iteration of the batch can be specified to ensure that selected candidates are the most beneficial to model learning. This new methodology is compared to B-EMCM using mean absolute error and root mean square error over ten iterations using benchmark machine learning data sets. Using multiple data sets and metrics across all methods, one of the variations of ABEMCM, that uses the max bound of the accumulated error (AB-EMCM Max), showed the best results for an adaptive batch approach. It achieved better root mean squared error (RMSE) and mean absolute error (MAE) than the other adaptive and nonadaptive batch methods while reaching the result in nearly the same number of iterations as the non-adaptive batch methods.

Share

COinS