University of Twente Student Theses

Login

Comparing supervised machine learning algorithms for client-specific care plans

Pistorius, C. (2023) Comparing supervised machine learning algorithms for client-specific care plans.

[img] PDF
1MB
Abstract:Creating client-specific care plans is a complex task where mistakes are easily made, especially by inexperienced caregivers. Supervised machine learning models can support these caregivers by suggesting relevant actions for client-specific care plans. In this research, four different algorithms are compared to determine whether supervised machine learning can provide actions based on scalar data of EHRs of clients. These algorithms suggest actions for clients with one of three illnesses: heart failure, dementia and diabetes. The BR-RF has the highest weighted F1-score, precision and recall on the test set. Furthermore, the created care plans of the BR-RF are compared to experienced caregivers’ care plans. The comparison shows that there is promise in the models, but that they are not performing well enough. Out of eight caregivers, five preferred the model’s care plans for heart failure clients, one selected the model’s care plan for a client with dementia and none chose the model’s care plan for clients with diabetes. The models cannot currently be implemented in a real-life situation based on the values of the performance metrics and the results of the comparison between a model’s and an experienced caregiver’s care plans.
Item Type:Essay (Master)
Clients:
Ecare, Enschede, Netherlands
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:31 mathematics, 50 technical science in general
Programme:Industrial Engineering and Management MSc (60029)
Link to this item:https://purl.utwente.nl/essays/97537
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page