Sensitivitätsanalyse

Sensitivitätsanalyse is a method used to determine how the uncertainty in the output of a mathematical model or computer program can be divided and attributed to different sources of uncertainty in its inputs. It is a crucial tool in risk assessment, decision-making, and scientific research, helping to understand which parameters have the most significant impact on a model's results.

The core idea behind Sensitivitätsanalyse is to systematically vary the input parameters of a model within their plausible ranges and observe the corresponding changes in the model's output. This process helps to identify the "sensitive" parameters, those whose variations lead to substantial shifts in the outcome. Understanding these sensitive parameters allows researchers and analysts to focus their efforts on obtaining more accurate data for them or to better understand the inherent uncertainties of the system being modeled.

There are various methods for performing Sensitivitätsanalyse. One common approach is local analysis, which examines the effect of small changes in input parameters around a specific point, often using derivatives. Global analysis, on the other hand, considers the entire range of possible input values and assesses the overall influence of each parameter. Techniques like variance-based methods, such as Sobol indices, are popular for global analysis as they quantify the contribution of each input parameter to the total variance of the model output.

The results of a Sensitivitätsanalyse can be presented in various ways, including sensitivity indices, tornado plots, or scatter plots. These visualizations help in communicating the importance of different factors and their impact on the model's predictions, thereby supporting more robust and informed conclusions.