Högskolan i Skövde

his.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • apa-cv
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Pre-bed-exit prediction using LSTM recurrent neural network
University of Skövde, School of Informatics.
2022 (English)Independent thesis Advanced level (degree of Master (One Year)), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

Continuous increase in fall incidents for the elderly has resulted in significant medical expenditure. As a result, there is a need for fall prevention. Fall detection system can be used to detect when a person is about to get out of bed such that early assistance can be provided. In this study, Long short-term memory neural network, a recurrent neural network capable of handling time series data, was proposed for predicting a pre-bed exit event using a simple four bed pressure sensor system. Still, pre-exit and exit states were classified based on the value from four pressure sensors, one placed under each leg of the bed. Sliding windowof 3, 5 and 10 seconds was used in assessing model performance. Different hyperparameter settings were applied in model tuning. A baseline mode, random forest, was used for comparison. To evaluate the performance, confusion matrix, accuracy, precision and recall were used. Overall accuracy of 95% was achieved when testing with near term data using LSTM model. Accuracy of random forest classifier was 90%. However, the accuracy of both models dropped when further away dates were used. The models were close in accuracy but made different mistakes.

Place, publisher, year, edition, pages
2022. , p. 45
National Category
Information Systems
Identifiers
URN: urn:nbn:se:his:diva-21535OAI: oai:DiVA.org:his-21535DiVA, id: diva2:1680033
Subject / course
Informationsteknologi
Educational program
Data Science - Master’s Programme
Supervisors
Examiners
Available from: 2022-07-02 Created: 2022-07-02 Last updated: 2022-07-02Bibliographically approved

Open Access in DiVA

No full text in DiVA

By organisation
School of Informatics
Information Systems

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 429 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • apa-cv
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf