Data envelope decision analysis

Submitted by Sukaina Bharwani | published 25th Mar 2011 | last updated 30th Mar 2011
Please note: content is older than 5 years

2D plot of DEA - this is used in business

2D plot of DEA - this is used in business

This idea is from the Data Envelope Analysis (DEA), but the purpose is very different. DEA is used to study the efficiency of each Decision Making Unit (DMU). For example, if an input and output is plotted in a 2D plot, it will be like the figure on the right hand. The observed sample 'B' is the most efficient for one measured output as it produces more output in relation to an input. So, the DMU becomes the basis to judge the efficiency of other samples., i.e. in this case the efficiency rate is measure by [Output of 'B' ] / [Output of other DMUs]. In this case, A's efficiency can be improved without adding extra input within the boundary of A, A1, A2.

The measurement of efficiency is irrelevant to the climate envelope, so I will not explain this purpose of DEA beyond this. The relevant concept of DEA is non-parametric approach to find the frontier or range of data. In other words, DEA does not discuss the trend in sampled data or the probability of achieving certain level of outcomes. Instead, DEA will talk about the feasibility or possibility of certain outcome by showing the range of observed outcomes. As this does not talk about the probability, does not talk about the distribution of outcomes either.

There are some applications of DEA to ranked voting systems, this can be more relevant in context of choosing adaptation options under climate envelope. We need more studies on this subject to decide if we would like to pursue this tool.

External links