Multi-Instance data sets represents problems where there is a many-to-one relationship between feature vectors and its output attribute. Usually, the output attribute defines the label of the bag where the instance is located.

When learning a classifier for this kind of data, the classifier should classify a new bag to a determinated class if one (or more) of its associated instances belongs to it.

Each data file has the following structure:

• @relation: Name of the data set
• @attribute: Description of an attribute (one for each attribute)
• @inputs: List with the names of the input attributes
• @output: Name of the output attribute
• @data: Starting tag of the data

The first attribute will always be the name of the bag which the instance belongs to. Also, it is required that every instance belonging to a given bag has the same value for its output attribute (its class).

The rest of the file contains all the examples belonging to the data set, expressed in comma sepparated values format.

We offer information about experimental studies using these data sets (result files, papers and more) in the Experimental studies with multi instance data sets section of the repository.