Use of Conventional Machine Learning to Optimize Deep Learning Hyper-parameters for NLP Labeling Tasks

Date
2020-01-07
Authors
Gu, Yang
Leroy, Gondy
Contributor
Advisor
Department
Instructor
Depositor
Speaker
Researcher
Consultant
Interviewer
Annotator
Journal Title
Journal ISSN
Volume Title
Publisher
Volume
Number/Issue
Starting Page
Ending Page
Alternative Title
Abstract
Deep learning delivers good performance in classification tasks, but is suboptimal with small and unbalanced datasets, which are common in many domains. To address this limitation, we use conventional machine learning, i.e., support vector machines (SVM) to tune deep learning hyper-parameters. We evaluated our approach using mental health electronic health records in which diagnostic criteria needed to extracted. A bidirectional Long Short-Term Memory network (BI-LSTM) could not learn the labels for the seven scarcest classes, but saw an increase in performance after training with optimal weights learned from tuning SVMs. With these customized class weights, the F1 scores for rare classes rose from 0 to values ranging from 18% to 57%. Overall, the BI-LSTM with SVM customized class weights achieved a micro-average of 47.1% for F1 across all classes, an improvement over the regular BI-LSTM’s 45.9%. The main contribution lies in avoiding null performance for rare classes.
Description
Keywords
Data, Text, and Web Mining for Business Analytics, deep learning, design science, machine learning, natural language processing
Citation
Extent
10 pages
Format
Geographic Location
Time Period
Related To
Proceedings of the 53rd Hawaii International Conference on System Sciences
Table of Contents
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International
Rights Holder
Local Contexts
Email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.