Enhancing DEA model : focus on time-controlled processes / Rasidah Mahdi
Data envelopment analysis (DEA) models have been applied to assess the relative efficiency of decision-making units, (DMUs). Here, the research problems focused on extending the application of DEA to incorporate time as one of the variables into the model formulation. The other focuses of this resea...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2009
|
| Subjects: | |
| Online Access: | https://ir.uitm.edu.my/id/eprint/5926/2/TP_RASIDAH%20MAHDI%20CS%2009_5%201.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Data envelopment analysis (DEA) models have been applied to assess the relative efficiency of decision-making units, (DMUs). Here, the research problems focused on extending the application of DEA to incorporate time as one of the variables into the model formulation. The other focuses of this research are to enhance the ranking approaches and to explore a method of reducing the subjectivity of obtaining the weight bounds. Weight bounds are an example of values used to impose weight restrictions in DEA models. To date, time has not being included as one of the variables in the DEA models formulation. This is needed when assessing the relative efficiency of processes with time controlled. The data from the clinical tests done at hospital laboratories, specifically at Seremban General Hospital were used to illustrate the extension of the application. Time was incorporated as one of the output variables into the DEA model formulated. This model was used to obtain the relative efficiency of the processes carried out. By incorporating time, the model discriminated the processes better. A Modified Cross Efficiency Matrix (MCEM) approach was introduced to enhance the two commonly ranking approaches, Andersen Petersen (AP) and Cross Efficiency Matrix (CEM). The MCEM approach introduced, based on the CEM approach, was found to give more consistent ranking results and found to be less sensitive to possible erroneous values. It was able to rank the DMU when AP approach produced the infeasible solutions. This result was illustrated by simulated data. |
|---|