Do Not Imitate, Reinforce: Iterative Classification via Belief Refinement
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917431733125120 |
|---|---|
| author | Kallel, Mahdi Tölle, Johannes Hendawy, Ahmed D'Eramo, Carlo |
| author_facet | Kallel, Mahdi Tölle, Johannes Hendawy, Ahmed D'Eramo, Carlo |
| contents | Standard supervised classification trains models to imitate the exact labels provided by a perfect oracle. This imitation happens in a single pass, restricting the model to a fixed compute budget even when inputs vary in complexity. Moreover, the rigid training objective forces the model to express absolute certainty on its training data, resulting in overconfident predictions during evaluation. We propose Reinforced Iterative Classification (RIC), which replaces the imitative objective with Reinforcement Learning (RL). RIC deploys a recurrent agent that iteratively updates a predictive distribution over classes, receiving reward for stepwise improvement in prediction quality. The value function provides a natural halting criterion by estimating the remaining scope for improvement. We prove that the iterative formulation recovers the same optimal predictions as cross-entropy while yielding an anytime classifier. On image classification benchmarks, RIC matches the accuracy of supervised baselines with improved calibration and learns to allocate computation adaptively across inputs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_22110 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Do Not Imitate, Reinforce: Iterative Classification via Belief Refinement Kallel, Mahdi Tölle, Johannes Hendawy, Ahmed D'Eramo, Carlo Machine Learning Standard supervised classification trains models to imitate the exact labels provided by a perfect oracle. This imitation happens in a single pass, restricting the model to a fixed compute budget even when inputs vary in complexity. Moreover, the rigid training objective forces the model to express absolute certainty on its training data, resulting in overconfident predictions during evaluation. We propose Reinforced Iterative Classification (RIC), which replaces the imitative objective with Reinforcement Learning (RL). RIC deploys a recurrent agent that iteratively updates a predictive distribution over classes, receiving reward for stepwise improvement in prediction quality. The value function provides a natural halting criterion by estimating the remaining scope for improvement. We prove that the iterative formulation recovers the same optimal predictions as cross-entropy while yielding an anytime classifier. On image classification benchmarks, RIC matches the accuracy of supervised baselines with improved calibration and learns to allocate computation adaptively across inputs. |
| title | Do Not Imitate, Reinforce: Iterative Classification via Belief Refinement |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2604.22110 |