Enhanced Multi-model Online Conformal Prediction
This work addresses computational bottlenecks in uncertainty quantification for online environments, offering an incremental improvement over prior multi-model approaches.
The paper tackles the problem of suboptimal prediction set size and high computational cost in multi-model online conformal prediction by developing a novel algorithm that uses bipartite graphs to select effective models at each time step, resulting in improved prediction set size and computational efficiency compared to existing techniques.
Conformal prediction is a framework for uncertainty quantification that constructs prediction sets for previously unseen data, guaranteeing coverage of the true label with a specified probability. However, the efficiency of these prediction sets, measured by their size, depends on the choice of the underlying learning model. Relying on a single fixed model may lead to suboptimal performance in online environments, as a single model may not consistently perform well across all time steps. To mitigate this, prior work has explored selecting a model from a set of candidates. However, this approach becomes computationally expensive as the number of candidate models increases. Moreover, poorly performing models in the set may also hinder the effectiveness. To tackle this challenge, this work develops a novel multi-model online conformal prediction algorithm that reduces computational complexity and improves prediction efficiency. At each time step, a bipartite graph is generated to identify a subset of effective models, from which a model is selected to construct the prediction set. Experiments demonstrate that our method outperforms existing multi-model conformal prediction techniques in terms of both prediction set size and computational efficiency.