We prove that every bounded function $f:\{-1,1\}^n\to[-1,1]$ of degree at most $d$ can be learned with $L_2$-accuracy $\varepsilon$ and confidence $1-\delta$ from $\log(\tfrac{n}{\delta})\,\varepsilon^{-d-1} C^{d^{3/2}\sqrt{\log d}}$ random queries, where $C>1$ is a universal finite constant.

翻译：我们证明,每个受约束的函数$f: ⁇ -1-1 ⁇ n\to[-1-1,1]美元,最多不超过1美元,可以用美元=2美元-准确性$\varepsilon$和信任值$1\delta$(tfrac{nhdelta})\\\\\,\varepsilon ⁇ _-d-1}C>d ⁇ 3/2 ⁇ sqrt}sqrt} d ⁇ $随机查询来学习,其中,$C>1是通用的有限常数。