arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1802.09419cs.LG
收藏

基于超网络的随机超参数优化

Stochastic Hyperparameter Optimization through Hypernetworks

Department of Computer Science, University of Toronto(多伦多大学计算机科学系)

Jonathan Lorraine, David Duvenaud

摘要:

机器学习模型通常通过在超参数优化内部嵌套优化模型权重来调整。我们给出了一种方法,将这种嵌套优化分解为权重和超参数的联合随机优化。我们的过程训练一个神经网络,以输出近似最优的权重作为超参数的函数。我们表明,我们的技术收敛到局部最优的权重和超参数足够大的超网络。我们比较这种方法标准的超参数优化策略,并证明其有效性调整数以千计的超参数。

英文摘要:

Machine learning models are often tuned by nesting optimization of model weights inside the optimization of hyperparameters. We give a method to collapse this nested optimization into joint stochastic optimization of weights and hyperparameters. Our process trains a neural network to output approximately optimal weights as a function of hyperparameters. We show that our technique converges to locally optimal weights and hyperparameters for sufficiently large hypernetworks. We compare this method to standard hyperparameter optimization strategies and demonstrate its effectiveness for tuning thousands of hyperparameters.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑