Combinatorial Characterizations of the Learnability of Conditional Distributions

概要

We study the PAC-learnability of function classes that map inputs to probability distributions over a finite space. The learner observes i.i.d. input-label pairs, with each label sampled from the conditional distribution of an unknown target hypothesis in the class. We measure learning through excess risk under a bounded proper loss that satisfies two mild assumptions. We introduce the \textbf{fat$_γ$Natarajan dimension}, which generalizes both fat$_γ$ and Natarajan dimension, and prove that its finiteness at every positive scale is necessary and sufficient for learnability. We also introduce a \textbf{fat$_γ$Graph} dimension extending the classical \gra dimension and establish inequalities comparing the two dimensions at different scales, leading to an equivalent characterization of learnability. Finally, we show that a distribution-valued function class is learnable if and only if each of its coordinate classes is learnable in the probabilistic concept model.

参考リンク

• Paper ↗

リンク先のライセンス・再現性・データ条件をご確認ください。自動分類や実装の目安は保証ではありません。

実装を始めるプロンプト

参考リンクを確認したうえで、次の内容をコピーして利用できます。

# 論文・記事
Combinatorial Characterizations of the Learnability of Conditional Distributions

# 概要
We study the PAC-learnability of function classes that map inputs to probability distributions over a finite space. The learner observes i.i.d.

# 原典
https://arxiv.org/abs/2610.31890

上記の手法を実装してください。実行前に依存関係とライセンスを確認してください。