我们研究了具有一般函数近似的部分可观察的MDP(POMDP)的外部评估(OPE)。现有的方法,例如顺序重要性采样估计器和拟合-Q评估,受POMDP中的地平线的诅咒。为了解决这个问题,我们通过引入将未来代理作为输入的未来依赖性值函数来开发一种新颖的无模型OPE方法。未来依赖性的价值函数在完全可观察的MDP中起着与经典价值函数相似的角色。我们为未来依赖性价值作为条件矩方程提供了一个新的Bellman方程,将历史记录代理用作仪器变量。我们进一步提出了一种最小值学习方法,以使用新的Bellman方程来学习未来依赖的价值函数。我们获得PAC结果,这意味着我们的OPE估计器是一致的,只要期货和历史包含有关潜在状态和Bellman完整性的足够信息。最后,我们将方法扩展到学习动力学,并在POMDP中建立我们的方法与众所周知的光谱学习方法之间的联系。
translated by 谷歌翻译
我们将自动辩护的机器学习的想法扩展到动态处理方案,并将其更普遍地扩展到嵌套功能。我们表明,可以根据递归riesz的代表表征嵌套平均回归的递归riesz代表来重新说明动态治疗方案的多重强大公式。然后,我们应用递归RIES代表估计学习算法,该学习算法估算偏低的校正,而无需表征校正术语的外观,例如,逆向概率加权术语的产物,如先前在双重稳健估计上所做的那样在动态状态中。我们的方法定义了一系列损失最小化问题的序列,其最小化是偏见校正的误解器,因此规避了解决辅助倾向模型的需求,并直接优化目标降低偏见校正的平均平方误差。我们为动态离散选择模型的估计提供了进一步的应用。
translated by 谷歌翻译
我们推出了一般,但简单,尖锐的界限,用于广泛的因果参数的省略可变偏置,可以被识别为结果的条件期望函数的线性功能。这些功能包括许多传统的因果推断研究中的调查目标,例如(加权)平均潜在结果,平均治疗效果(包括亚组效应,例如对处理的效果),(加权)平均值来自协变态分布的转变的衍生品和政策影响 - 所有是一般的非参数因果模型。我们的建设依赖于目标功能的riesz-frechet表示。具体而言,我们展示了偏差的绑定如何仅取决于潜在变量在结果中创建的附加变型以及用于感兴趣的参数的RIESZ代表。此外,在许多重要病例中(例如,部分线性模型中的平均治疗效果,或在具有二元处理的不可分配模型中),所示的界定依赖于两个易于解释的数量:非参数部分$ r ^ 2 $(Pearson的相关性与治疗和结果的未观察变量的比例“。因此,对省略变量的最大解释力(在解释处理和结果变化时)的简单合理性判断足以将整体界限放置在偏置的尺寸上。最后,利用脱叠机器学习,我们提供灵活有效的统计推理方法,以估计从观察到的分布识别的界限的组件。
translated by 谷歌翻译
感兴趣的许多因果和政策效应都是由高维或非参数回归函数的线性功能定义的。 $ \ sqrt {n} $ - 对目标对象的一致且渐近地正常估计需要偏见,以减少正则化和/或模型选择对感兴趣对象的影响。通常,通过将校正项添加到功能的插件估计器中来实现,从而导致属性,例如半参数效率,双重鲁棒性和Neyman正交性。我们基于自动学习使用神经网和随机森林的Riesz表示的自动偏差程序。我们的方法仅依赖于黑框评估Oracle访问线性功能,并且不需要其分析形式的知识。我们提出了一种多任务神经网络偏见方法,具有随机梯度下降最小化的Riesz代表和回归损失,同时共享这两个函数的表示层。我们还提出了一种随机森林方法,该方法了解Riesz函数的局部线性表示。即使我们的方法适用于任意功能,我们在实验上发现它的性能与Shi等人的最先进的神经网状算法相比。 (2019)对于平均治疗效果功能的情况。我们还使用汽油需求的汽油价格变化的半合成数据来评估我们的方法,即通过连续处理估算平均边缘效应的问题。
translated by 谷歌翻译
Doubleml是一个开源的Python库,实现Chernozhukov等人的双层机器学习框架。(2018)适用于各种因果模型。它包含有效统计推断对因果参数的有效推断的功能基于机器学习方法。面向对象的Doublem实施在型号规格方面提供了很高的灵活性,并使其轻松伸展。该包在麻省理工学院许可下分发,并依赖于科学Python生态系统的核心库:Scikit-Learn,Numpy,Pandas,Scipy,StatsModels和Joblib。源代码,文档和广泛的用户指南可以在https://github.com/doubleml/doubleml-for -py和https://docs.doubleml.org找到。
translated by 谷歌翻译
R包Doubleml实现了Chernozhukov等人的双重/辩护机器学习框架。 (2018)。它提供了基于机器学习方法的因果模型中估计参数的功能。双机器学习框架由三个关键成分组成:Neyman正交性,高质量的机器学习估计和样品拆分。可以通过MLR3生态系统中可用的各种最新机器学习方法来执行滋扰组件的估计。 Doubleml使得可以在各种因果模型中进行推断,包括部分线性和交互式回归模型及其扩展到仪器变量估计。 Doubleml的面向对象的实现为模型规范具有很高的灵活性,并使其易于扩展。本文是对双机器学习框架和R软件包DOUBLEML的介绍。在具有模拟和真实数据集的可再现代码示例中,我们演示了Doubleml用户如何基于机器学习方法执行有效的推断。
translated by 谷歌翻译
本文涉及根N的可行性和手段,始终估算高维,大约稀疏回归的线性,均方连续功能。这些对象包括各种有趣的参数,例如回归系数,平均衍生物和平均治疗效果。我们给出了回归斜率和平均导数的估计量的收敛速率的下限,并发现这些界限大大比低维,半参数设置大。我们还提供了依据的机器学习者,这些学习者在最小的稀疏条件或速率双重鲁棒性下是一致的。这些估计值对在先前已知的更一般条件下保持root-n一致的现有估计值有所改善。
translated by 谷歌翻译
This paper provides estimation and inference methods for a conditional average treatment effects (CATE) characterized by a high-dimensional parameter in both homogeneous cross-sectional and unit-heterogeneous dynamic panel data settings. In our leading example, we model CATE by interacting the base treatment variable with explanatory variables. The first step of our procedure is orthogonalization, where we partial out the controls and unit effects from the outcome and the base treatment and take the cross-fitted residuals. This step uses a novel generic cross-fitting method we design for weakly dependent time series and panel data. This method "leaves out the neighbors" when fitting nuisance components, and we theoretically power it by using Strassen's coupling. As a result, we can rely on any modern machine learning method in the first step, provided it learns the residuals well enough. Second, we construct an orthogonal (or residual) learner of CATE -- the Lasso CATE -- that regresses the outcome residual on the vector of interactions of the residualized treatment with explanatory variables. If the complexity of CATE function is simpler than that of the first-stage regression, the orthogonal learner converges faster than the single-stage regression-based learner. Third, we perform simultaneous inference on parameters of the CATE function using debiasing. We also can use ordinary least squares in the last two steps when CATE is low-dimensional. In heterogeneous panel data settings, we model the unobserved unit heterogeneity as a weakly sparse deviation from Mundlak (1978)'s model of correlated unit effects as a linear function of time-invariant covariates and make use of L1-penalization to estimate these models. We demonstrate our methods by estimating price elasticities of groceries based on scanner data. We note that our results are new even for the cross-sectional (i.i.d) case.
translated by 谷歌翻译
Selecting the number of topics in LDA models is considered to be a difficult task, for which alternative approaches have been proposed. The performance of the recently developed singular Bayesian information criterion (sBIC) is evaluated and compared to the performance of alternative model selection criteria. The sBIC is a generalization of the standard BIC that can be implemented to singular statistical models. The comparison is based on Monte Carlo simulations and carried out for several alternative settings, varying with respect to the number of topics, the number of documents and the size of documents in the corpora. Performance is measured using different criteria which take into account the correct number of topics, but also whether the relevant topics from the DGPs are identified. Practical recommendations for LDA model selection in applications are derived.
translated by 谷歌翻译
Applying deep learning concepts from image detection and graph theory has greatly advanced protein-ligand binding affinity prediction, a challenge with enormous ramifications for both drug discovery and protein engineering. We build upon these advances by designing a novel deep learning architecture consisting of a 3-dimensional convolutional neural network utilizing channel-wise attention and two graph convolutional networks utilizing attention-based aggregation of node features. HAC-Net (Hybrid Attention-Based Convolutional Neural Network) obtains state-of-the-art results on the PDBbind v.2016 core set, the most widely recognized benchmark in the field. We extensively assess the generalizability of our model using multiple train-test splits, each of which maximizes differences between either protein structures, protein sequences, or ligand extended-connectivity fingerprints. Furthermore, we perform 10-fold cross-validation with a similarity cutoff between SMILES strings of ligands in the training and test sets, and also evaluate the performance of HAC-Net on lower-quality data. We envision that this model can be extended to a broad range of supervised learning problems related to structure-based biomolecular property prediction. All of our software is available as open source at https://github.com/gregory-kyro/HAC-Net/.
translated by 谷歌翻译