PyTorchKR Papers

최신 논문

총 3,318편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1807.10100

Two-Step Estimation and Inference with Possibly Many Included Covariates

We study the implications of including many covariates in a first-step estimate entering a two-step estimation procedure. We find that a first order bias emerges when the number of \textit{included} covariates is "large" relative to the square-root of sample size, rendering standard inference procedures invalid. We show that the jackknife is able to estimate this "many covariates" bias consistently, thereby delivering a new automatic bias-corrected two-step point estimator. The jackknife also consistently estimates the standard error of the original two-step point estimator. For inference, we develop a valid post-bias-correction bootstrap approximation that accounts for the additional variability introduced by the jackknife bias-correction. We find that the jackknife bias-corrected point estimator and the bootstrap post-bias-correction inference perform excellent in simulations, offering important improvements over conventional two-step point estimators and inference procedures, which are not robust to including many covariates. We apply our results to an array of distinct treatment effect, policy evaluation, and other applied microeconomics settings. In particular, we discuss production function and marginal treatment effect estimation in detail.

arXiv:1807.06461

Centralized Scheduling Strategies for Cooperative HARQ Retransmissions in Multi-Source Multi-Relay Wireless Networks

In this paper, we investigate centralized scheduling strategies for cooperative incremental redundancy retransmissions in the slow-fading half-duplex multiple access multiple relay channel. Time Division Multiple Access is assumed for the sources and the relays. Sources transmit successively in time slots for the first phase. The second phase consists of a limited number of time slots for retransmissions. In each time slot, the destination schedules a node (being a relay or a source) to retransmit, conditional on the knowledge of the correctly decoded source sets of each node (which is itself for a source). A scheduled relay retransmission uses Joint Network and Channel Coding on its correctly decoded source messages (cooperative retransmission). Several node selection strategies are proposed based on the maximization of the long-term aggregate throughput under a given constraint of fairness. Monte-Carlo simulations show that these strategies outperform the state of the art one based on the minimization of the probability of the common outage event after each time-slot. Moreover, the long-term aggregate throughput reached with these strategies is close to the upper-bound, calculated by the exhaustive search approach. The same conclusion remains valid for both symmetric and asymmetric source rate scenarios.

arXiv:1807.06338

Limit Theorems for Factor Models

The paper establishes the central limit theorems and proposes how to perform valid inference in factor models. We consider a setting where many counties/regions/assets are observed for many time periods, and when estimation of a global parameter includes aggregation of a cross-section of heterogeneous micro-parameters estimated separately for each entity. The central limit theorem applies for quantities involving both cross-sectional and time series aggregation, as well as for quadratic forms in time-aggregated errors. The paper studies the conditions when one can consistently estimate the asymptotic variance, and proposes a bootstrap scheme for cases when one cannot. A small simulation study illustrates performance of the asymptotic and bootstrap procedures. The results are useful for making inferences in two-step estimation procedures related to factor models, as well as in other related contexts. Our treatment avoids structural modeling of cross-sectional dependence but imposes time-series independence.

arXiv:1807.04788

Systematic evaluation of the population-level effects of alternative treatment strategies on the basic reproduction number

An approach to estimate the influence of the treatment-type controls on the basic reproduction number, R 0 , is proposed and elaborated. The presented approach allows one to estimate the effect of a given treatment strategy or to compare a number of different treatment strategies on the basic reproduction number. All our results are valid for sufficiently small values of the control. However, in many cases it is possible to extend this analysis to larger values of the control as was illustrated by examples.

arXiv:1807.03257

Data Efficient Lithography Modeling with Transfer Learning and Active Data Selection

Lithography simulation is one of the key steps in physical verification, enabled by the substantial optical and resist models. A resist model bridges the aerial image simulation to printed patterns. While the effectiveness of learning-based solutions for resist modeling has been demonstrated, they are considerably data-demanding. Meanwhile, a set of manufactured data for a specific lithography configuration is only valid for the training of one single model, indicating low data efficiency. Due to the complexity of the manufacturing process, obtaining enough data for acceptable accuracy becomes very expensive in terms of both time and cost, especially during the evolution of technology generations when the design space is intensively explored. In this work, we propose a new resist modeling framework for contact layers, utilizing existing data from old technology nodes and active selection of data in a target technology node, to reduce the amount of data required from the target lithography configuration. Our framework based on transfer learning and active learning techniques is effective within a competitive range of accuracy, i.e., 3-10X reduction on the amount of training data with comparable accuracy to the state-of-the-art learning approach.

arXiv:1806.09304

Semiparametrically Point-Optimal Hybrid Rank Tests for Unit Roots

We propose a new class of unit root tests that exploits invariance properties in the Locally Asymptotically Brownian Functional limit experiment associated to the unit root model. The invariance structures naturally suggest tests that are based on the ranks of the increments of the observations, their average, and an assumed reference density for the innovations. The tests are semiparametric in the sense that they are valid, i.e., have the correct (asymptotic) size, irrespective of the true innovation density. For a correctly specified reference density, our test is point-optimal and nearly efficient. For arbitrary reference densities, we establish a Chernoff-Savage type result, i.e., our test performs as well as commonly used tests under Gaussian innovations but has improved power under other, e.g., fat-tailed or skewed, innovation distributions. To avoid nonparametric estimation, we propose a simplified version of our test that exhibits the same asymptotic properties, except for the Chernoff-Savage result that we are only able to demonstrate by means of simulations.

arXiv:1805.10677

On a stochastic model of epidemic spread with an application to competing infections

A simple, but ``classical``, stochastic model for epidemic spread in a finite, but large, population is studied. The progress of the epidemic can be divided into three different phases that requires different tools to analyse. Initially the process is approximated by a branching process. It is discussed for how long time this approximation is valid. When a non-negligible proportion of the population is already infected the process can be studied using differential equations. In a final phase the spread will fade out. The results are used to investigate what happens if two strains of infectious agents, with different potential for spread, are simultaneously introduced in a totally susceptible population. It is assumed that an infection causes immunity, and that a person can only be infected by one strain. The two epidemics will initially develop approximately as independent branching processes. However, if both strains causes large epidemics they will, due to immunity, eventually interact. We will mainly be interested in the final outcome of the spread, i.e., how large proportion of the population is infected by the different strains.

arXiv:1804.03349

Inference on Local Average Treatment Effects for Misclassified Treatment

We develop point-identification for the local average treatment effect when the binary treatment contains a measurement error. The standard instrumental variable estimator is inconsistent for the parameter since the measurement error is non-classical by construction. We correct the problem by identifying the distribution of the measurement error based on the use of an exogenous variable that can even be a binary covariate. The moment conditions derived from the identification lead to generalized method of moments estimation with asymptotically valid inferences. Monte Carlo simulations and an empirical illustration demonstrate the usefulness of the proposed procedure.

arXiv:1804.00232

Continuous Record Laplace-based Inference about the Break Date in Structural Change Models

Building upon the continuous record asymptotic framework recently introduced by Casini and Perron (2018a) for inference in structural change models, we propose a Laplace-based (Quasi-Bayes) procedure for the construction of the estimate and confidence set for the date of a structural change. It is defined by an integration rather than an optimization-based method. A transformation of the least-squares criterion function is evaluated in order to derive a proper distribution, referred to as the Quasi-posterior. For a given choice of a loss function, the Laplace-type estimator is the minimizer of the expected risk with the expectation taken under the Quasi-posterior. Besides providing an alternative estimate that is more precise|lower mean absolute error (MAE) and lower root-mean squared error (RMSE)|than the usual least-squares one, the Quasi-posterior distribution can be used to construct asymptotically valid inference using the concept of Highest Density Region. The resulting Laplace-based inferential procedure is shown to have lower MAE and RMSE, and the confidence sets strike the best balance between empirical coverage rates and average lengths of the confidence sets relative to traditional long-span methods, whether the break size is small or large.

arXiv:1803.09452

Panel Data Analysis with Heterogeneous Dynamics

This paper proposes a model-free approach to analyze panel data with heterogeneous dynamic structures across observational units. We first compute the sample mean, autocovariances, and autocorrelations for each unit, and then estimate the parameters of interest based on their empirical distributions. We then investigate the asymptotic properties of our estimators using double asymptotics and propose split-panel jackknife bias correction and inference based on the cross-sectional bootstrap. We illustrate the usefulness of our procedures by studying the deviation dynamics of the law of one price. Monte Carlo simulations confirm that the proposed bias correction is effective and yields valid inference in small samples.

arXiv:1803.09020

Schooling Choice, Labour Market Matching, and Wages

We develop inference for a two-sided matching model where the characteristics of agents on one side of the market are endogenous due to pre-matching investments. The model can be used to measure the impact of frictions in labour markets using a single cross-section of matched employer-employee data. The observed matching of workers to firms is the outcome of a discrete, two-sided matching process where firms with heterogeneous preferences over education sequentially choose workers according to an index correlated with worker preferences over firms. The distribution of education arises in equilibrium from a Bayesian game: workers, knowing the distribution of worker and firm types, invest in education prior to the matching process. Although the observed matching exhibits strong cross-sectional dependence due to the matching process, we propose an asymptotically valid inference procedure that combines discrete choice methods with simulation.

arXiv:1803.09015

Difference-in-Differences with Multiple Time Periods

In this article, we consider identification, estimation, and inference procedures for treatment effect parameters using Difference-in-Differences (DID) with (i) multiple time periods, (ii) variation in treatment timing, and (iii) when the "parallel trends assumption" holds potentially only after conditioning on observed covariates. We show that a family of causal effect parameters are identified in staggered DID setups, even if differences in observed characteristics create non-parallel outcome dynamics between groups. Our identification results allow one to use outcome regression, inverse probability weighting, or doubly-robust estimands. We also propose different aggregation schemes that can be used to highlight treatment effect heterogeneity across different dimensions as well as to summarize the overall effect of participating in the treatment. We establish the asymptotic properties of the proposed estimators and prove the validity of a computationally convenient bootstrap procedure to conduct asymptotically valid simultaneous (instead of pointwise) inference. Finally, we illustrate the relevance of our proposed tools by analyzing the effect of the minimum wage on teen employment from 2001--2007. Open-source software is available for implementing the proposed methods.

arXiv:1803.07951

Testing Continuity of a Density via g-order statistics in the Regression Discontinuity Design

In the regression discontinuity design (RDD), it is common practice to assess the credibility of the design by testing the continuity of the density of the running variable at the cut-off, e.g., McCrary (2008). In this paper we propose an approximate sign test for continuity of a density at a point based on the so-called g-order statistics, and study its properties under two complementary asymptotic frameworks. In the first asymptotic framework, the number q of observations local to the cut-off is fixed as the sample size n diverges to infinity, while in the second framework q diverges to infinity slowly as n diverges to infinity. Under both of these frameworks, we show that the test we propose is asymptotically valid in the sense that it has limiting rejection probability under the null hypothesis not exceeding the nominal level. More importantly, the test is easy to implement, asymptotically valid under weaker conditions than those used by competing methods, and exhibits finite sample validity under stronger conditions than those needed for its asymptotic validity. In a simulation study, we find that the approximate sign test provides good control of the rejection probability under the null hypothesis while remaining competitive under the alternative hypothesis. We finally apply our test to the design in Lee (2008), a well-known application of the RDD to study incumbency advantage.

arXiv:1801.07988

Understanding news story chains using information retrieval and network clustering techniques

Content analysis of news stories (whether manual or automatic) is a cornerstone of the communication studies field. However, much research is conducted at the level of individual news articles, despite the fact that news events (especially significant ones) are frequently presented as "stories" by news outlets: chains of connected articles covering the same event from different angles. These stories are theoretically highly important in terms of increasing public recall of news items and enhancing the agenda-setting power of the press. Yet thus far, the field has lacked an efficient method for detecting groups of articles which form stories in a way that enables their analysis. In this work, we present a novel, automated method for identifying linked news stories from within a corpus of articles. This method makes use of techniques drawn from the field of information retrieval to identify textual closeness of pairs of articles, and then clustering techniques taken from the field of network analysis to group these articles into stories. We demonstrate the application of the method to a corpus of 61,864 articles, and show how it can efficiently identify valid story clusters within the corpus. We use the results to make observations about the prevalence and dynamics of stories within the UK news media, showing that more than 50% of news production takes place within stories.

arXiv:1801.03627

Applying Vector Space Model (VSM) Techniques in Information Retrieval for Arabic Language

Information Retrieval (IR) allows the storage, management, processing and retrieval of information, documents, websites, etc. Building an IR system for any language is imperative. This is evident through the massive conducted efforts to build IR systems using any of its models that are valid for certain languages. This report presents an implementation for a core IR technique which is Vector Space Model (VSM). We have chosen VSM model for our project since it is a term weighting scheme, and the retrieved documents could be sorted according to their relevancy degree. One other significant feature for such technique is the ability to get a relevance feedback from the users of the system; users can judge whether the retrieved document is relative to their need or not. The developed system has been validated through building an Arabic IR website using server side scripting. The experiments verifies the effectiveness of our system to apply all techniques of vector space model and valid over Arabic language.

arXiv:1712.08102

Simultaneous Confidence Intervals for High-dimensional Linear Models with Many Endogenous Variables

High-dimensional linear models with endogenous variables play an increasingly important role in recent econometric literature. In this work we allow for models with many endogenous variables and many instrument variables to achieve identification. Because of the high-dimensionality in the second stage, constructing honest confidence regions with asymptotically correct coverage is non-trivial. Our main contribution is to propose estimators and confidence regions that would achieve that. The approach relies on moment conditions that have an additional orthogonal property with respect to nuisance parameters. Moreover, estimation of high-dimension nuisance parameters is carried out via new pivotal procedures. In order to achieve simultaneously valid confidence regions we use a multiplier bootstrap procedure to compute critical values and establish its validity.

arXiv:1712.08081

Overview of the Triple Scoring Task at the WSDM Cup 2017

This paper provides an overview of the triple scoring task at the WSDM Cup 2017, including a description of the task and the dataset, an overview of the participating teams and their results, and a brief account of the methods employed. In a nutshell, the task was to compute relevance scores for knowledge-base triples from relations, where such scores make sense. Due to the way the ground truth was constructed, scores were required to be integers from the range 0..7. For example, reasonable scores for the triples "Tim Burton profession Director" and "Tim Burton profession Actor" would be 7 and 2, respectively, because Tim Burton is well-known as a director, but he acted only in a few lesser known movies. The triple scoring task attracted considerable interest, with 52 initial registrations and 21 teams who submitted a valid run before the deadline. The winning team achieved an accuracy of 87%, that is, for that fraction of the triples from the test set (which was revealed only after the deadline) the difference to the score from the ground truth was at most 2. The best result for the average difference from the test set scores was 1.50.

arXiv:1712.07522

Cointegration in functional autoregressive processes

This paper defines the class of $\mathcal{H}$-valued autoregressive (AR) processes with a unit root of finite type, where $\mathcal{H}$ is an infinite dimensional separable Hilbert space, and derives a generalization of the Granger-Johansen Representation Theorem valid for any integration order $d=1,2,\dots$. An existence theorem shows that the solution of an AR with a unit root of finite type is necessarily integrated of some finite integer $d$ and displays a common trends representation with a finite number of common stochastic trends of the type of (cumulated) bilateral random walks and an infinite dimensional cointegrating space. A characterization theorem clarifies the connections between the structure of the AR operators and $(i)$ the order of integration, $(ii)$ the structure of the attractor space and the cointegrating space, $(iii)$ the expression of the cointegrating relations, and $(iv)$ the Triangular representation of the process. Except for the fact that the number of cointegrating relations that are integrated of order 0 is infinite, the representation of $\mathcal{H}$-valued ARs with a unit root of finite type coincides with that of usual finite dimensional VARs, which corresponds to the special case $\mathcal{H}=\mathbb{R}^p$.

arXiv:1707.00610

Option Pricing under Fast-varying and Rough Stochastic Volatility

Recent empirical studies suggest that the volatilities associated with financial time series exhibit short-range correlations. This entails that the volatility process is very rough and its autocorrelation exhibits sharp decay at the origin. Another classic stylistic feature often assumed for the volatility is that it is mean reverting. In this paper it is shown that the price impact of a rapidly mean reverting rough volatility model coincides with that associated with fast mean reverting Markov stochastic volatility models. This reconciles the empirical observation of rough volatility paths with the good fit of the implied volatility surface to models of fast mean reverting Markov volatilities. Moreover, the result conforms with recent numerical results regarding rough stochastic volatility models. It extends the scope of models for which the asymptotic results of fast mean reverting Markov volatilities are valid. The paper concludes with a general discussion of fractional volatility asymptotics and their interrelation. The regimes discussed there include fast and slow volatility factors with strong or small volatility fluctuations and with the limits not commuting in general. The notion of a characteristic term structure exponent is introduced, this exponent governs the implied volatility term structure in the various asymptotic regimes.

arXiv:1706.09067

Structured Recommendation

Current recommender systems largely focus on static, unstructured content. In many scenarios, we would like to recommend content that has structure, such as a trajectory of points-of-interests in a city, or a playlist of songs. Dubbed Structured Recommendation, this problem differs from the typical structured prediction problem in that there are multiple correct answers for a given input. Motivated by trajectory recommendation, we focus on sequential structures but in contrast to classical Viterbi decoding we require that valid predictions are sequences with no repeated elements. We propose an approach to sequence recommendation based on the structured support vector machine. For prediction, we modify the inference procedure to avoid predicting loops in the sequence. For training, we modify the objective function to account for the existence of multiple ground truths for a given input. We also modify the loss-augmented inference procedure to exclude the known ground truths. Experiments on real-world trajectory recommendation datasets show the benefits of our approach over existing, non-structured recommendation approaches.