Benefits of machine learning in operational management systems in the social and healthcare sectors
Year: 2024
Type: dissertation
Abstract: This master thesis aimed to provide a practical understanding of how machine learning (ML) can offer advantages in operational management systems, particularly in Social- and Healthcare districts. The research was conducted as a case study with two research phases. The first part of the case study involved descriptive integrative literature research, which clarified the stage of AI usage in Finland's healthcare system and its regulatory landscape. Practical examples from healthcare and other industries were researched to enhance understanding of existing projects and potential use cases for supervised machine learning. The methodology of modern machine learning projects was also explored to provide insights into delivering a machine learning project in a chosen area. The second part of the case study focused on how the Machine Learning capability of the Solution Business Manager (SBM) could support operational management systems. The research included a pilot demonstration of SMB machine learning classification capability to determine its technical functionality and limitations. \nThe main result of this study was that Finland is still very far from the ideal way of utilizing machine learning in predictive models to find diagnoses and preventive care in the social and healthcare districts. The reason is that the regulations are stringent and deficient in artificial intelligence usage. On the other hand, there are many possibilities to use machine learning in this sector when considering all the supporting areas that are not utilizing patient data. Some to mention are management tasks, logistics improvements, predictive maintenance, operation improvements like workflows or patient coordination based on more complicated rules, data classifications for reporting purposes, and future big data AI solutions. Synthetic data could solve some of the issues and help start research work in the future. The result of SBM´s first ML solution was that the technical functionalities are easy to use and are fast to build in existing solutions, which delivers cost advantages compared to external solutions. If not, a new solution can be built for new data prediction projects, but the data structure and requirements must be considered. The SBM solution is ideal for workflow improvements, calculations, process item directions based on complicated rules, many fields, or automatic classification of items based on long descriptions.
Source: LUTPub (LUT University)
Author: Huttunen, Jenni
Institution: Lappeenranta-Lahti University of Technology
Language: English
Subfield: Health Informatics
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7026937418
