Aggregation in conformal e-classification
For researchers in conformal prediction, this provides simpler aggregation methods that retain validity, though the improvements are incremental.
The paper studies aggregation of conformal e-predictors, focusing on cross-conformal e-prediction and its simpler modifications, showing they maintain validity while improving efficiency.
Aggregating conformal predictors is a standard way of balancing their predictive and computational efficiency while retaining their validity, at least approximately. An important advantage of conformal e-predictors is that they are easier to aggregate without sacrificing their validity. This paper studies experimentally cross-conformal e-prediction, which is an existing method of aggregating conformal e-predictors, and its modifications that are conceptually simpler and more flexible.