Forecasting healthcare expenditure using ConvLSTM

Thumbnail Image
Date
2023-05
Authors
Lawande, Saurabh Nitin
Major Professor
Quinn, Christopher
Advisor
Committee Member
Bao, Forrest Sheng
Journal Title
Journal ISSN
Volume Title
Publisher
Altmetrics
Abstract
There have been multiple methods proposed both neural and statistical, in literature, to help predict healthcare expenditure. The main goal of this work was to evaluate the neural techniques and use them to predict the average weekly expenditure by patients on medications. In this paper, autoregressive integrated moving averages and ConvLSTM models were used in forecasting. These models were carefully calibrated and used to predict healthcare expenditure. From observing the results, it was noted that the ConvLSTM clearly outperformed the ARIMA model in predicting the outcome for both the medications
Series Number
Journal Issue
Is Version Of
Versions
Series
Academic or Administrative Unit
Type
creative component
Comments
Rights Statement
Copyright
2023
Funding
Supplemental Resources
Source