Name
Abstract | ||
|---|---|---|
Reliability is an important characteristic of any system. Healthcare systems are typical examples of such systems. In reliability engineering, such systems are considered as complex, inhomogeneous, and uncertain, and require special mathematical representations. The structure function is a suitable model representing real systems. Methods of system reliability evaluation based on the structure function are well established but deterministic. This restricts its use for uncertain or incomplete data. A structure function can be created only for system in which correlations of all components are indicated and all component states are known. In this paper, a new method for structure function construction is proposed. Incomplete data is analysed using Fuzzy Decision Trees (FDTs), where input and output attributes are interpreted as component states and values of the structure function, respectively. This method is applied to reliability analysis of healthcare system. For illustration, we considered the system laparoscopic surgery that has 4 components and 36 state vectors. In addition we evaluate proposed method by 3 benchmarku0027s systems with 243, 108, and 512 state vectors, respectively. Two of these benchmarks have 5 components and one has 4 components. Uncertainty is simulated by randomly deleting between 5% and 90% of all state vectors before constructing the structure function. With 50% of deleted stages, the error rate is below 0.2% for all three systems. We conclude that FDT-based reliability analysis is applicable for incomplete data in medical systems, too. |
Year | Venue | Field |
|---|---|---|
2016 | FedCSIS | Data mining,Mathematical optimization,Computer science,Word error rate,Medical systems,Input/output,Artificial intelligence,Structure function,Healthcare system,Real systems,Fuzzy decision tree,Machine learning |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
0 | 4 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Elena N. Zaitseva | 1 | 53 | 14.38 |
| Vitaly G. Levashenko | 2 | 39 | 12.90 |
| Miroslav Kvassay | 3 | 9 | 8.85 |
| Thomas Martin Deserno | 4 | 530 | 67.07 |