arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1709.00291math.STmath.OCstat.MLstat.TH
收藏

随机梯度搜索的渐近偏差

Asymptotic Bias of Stochastic Gradient Search

Vladislav B. Tadic, Arnaud Doucet

摘要:

分析了带有偏梯度估计的随机梯度算法的渐近性态。依靠基于动态系统理论(链递归)和微分几何(Yomdin定理和Lojasiewicz不等式)的参数,得到了这样的算法产生的迭代的渐近偏差的紧界.所得到的结果保持在温和的条件下,并涵盖了广泛的一类高维非线性算法。利用这些结果,研究了策略梯度(强化)学习和自适应群体蒙特卡罗抽样的渐近性质。在此基础上,分析了隐马尔可夫模型中递推极大分裂似然估计的渐近性态。

英文摘要:

The asymptotic behavior of the stochastic gradient algorithm with a biased gradient estimator is analyzed. Relying on arguments based on the dynamic system theory (chain-recurrence) and the differential geometry (Yomdin theorem and Lojasiewicz inequality), tight bounds on the asymptotic bias of the iterates generated by such an algorithm are derived. The obtained results hold under mild conditions and cover a broad class of high-dimensional nonlinear algorithms. Using these results, the asymptotic properties of the policy-gradient (reinforcement) learning and adaptive population Monte Carlo sampling are studied. Relying on the same results, the asymptotic behavior of the recursive maximum split-likelihood estimation in hidden Markov models is analyzed, too.

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。

↑