Recommendation of a suitable XAI method

Insert values for getting a recommendation

In order to be able to propose the most appropriate XAI method for your ML model, you first of all have to provide some properties of your ML model and its underlying data. Additionally you can add personal preferences, for example regarding the method's explanation type or it's usage.

Model specific input parameters

shortlong

Data specific input parameters

The suitability of an XAI method highly depends on the nature of the training data. Therefore, some properties of these must be obtained and considered for a recommendation. Several options are available for the specification: An approximate specification via the tab "Approximate input" is usually already sufficient. However, if the required parameters are already available, you can specify them in the "Exact input" tab. Unfortunately, the option of automatic data analysis is not yet available.

Approximate input

negligiblehigh
poorgood
fewmany

User preferences regarding XAI method

doesn't matterhigh
fastdoesn't matter