中文说明：源代码的文件"功能进行哈希处理的大型规模多任务学习"。实证证据表明哈希是维数约简和实用的非参数估计的有效策略。我们提供指数尾边界有限元-真正进行哈希处理和显示随机子空间之间的相互作用是高概率可忽略不计。我们展示的可行性这种方法与实验结果为一个新的用例 — — 多任务学习与数以十万计的任务。
English Description：
Source code file & quot; function for hashing large-scale multi task learning & quot;. Empirical evidence shows that hash is an effective strategy for dimension reduction and practical nonparametric estimation. We provide exponential tail boundary finite element - true hashing and show that the interaction between random subspaces is highly probabilistic and negligible. We demonstrate the feasibility of this approach with experimental results for a new use case - multi task learning with hundreds of thousands of tasks.