Please use this identifier to cite or link to this item: http://hdl.handle.net/10125/63867

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

File Size Format  
0102.pdf 716.38 kB Adobe PDF View/Open

Item Summary

Title:Use of Conventional Machine Learning to Optimize Deep Learning Hyper-parameters for NLP Labeling Tasks
Authors:Gu, Yang
Leroy, Gondy
Keywords:Data, Text, and Web Mining for Business Analytics
deep learning
design science
machine learning
natural language processing
Date Issued:07 Jan 2020
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.
Pages/Duration:10 pages
URI:http://hdl.handle.net/10125/63867
ISBN:978-0-9981331-3-3
DOI:10.24251/HICSS.2020.128
Rights:Attribution-NonCommercial-NoDerivatives 4.0 International
https://creativecommons.org/licenses/by-nc-nd/4.0/
Appears in Collections: Data, Text, and Web Mining for Business Analytics


Please email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.

This item is licensed under a Creative Commons License Creative Commons