The motivation of my research is to combine models learnt by agents that have partial views of all the data available for a learning tasks. For example, local consumer models learnt from individual supermarkets might be combined to give a national model.

I will introduce a new technique for combining multiple learned models. This technique results in a single comprehensible model. This is to be contrasted with current methods that typically combine models by voting. The core of the technique, the DAGGER (Disjoint Aggregation using Example Reduction) algorithm selects examples which provide evidence for each decision region within each local model. A single model is then learned from the union of these selected examples.
 

Return to the seminar schedule