DOCSLIB.ORG

Learning Entangled Single-Sample Distributions Via Iterative Trimming

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Learning Entangled Single-Sample Distributions Via Iterative Trimming Learning Entangled Single-Sample Distributions via Iterative Trimming Hui Yuan Yingyu Liang Department of Statistics and Finance Department of Computer Sciences University of Science and Technology of China University of Wisconsin-Madison Abstract have n data points from n different distributions with a common mean but different variances (the mean and all the variances are unknown), and our goal is to es- In the setting of entangled single-sample dis- timate the mean. tributions, the goal is to estimate some com- mon parameter shared by a family of distri- This setting is motivated for both theoretical and prac- butions, given one single sample from each tical reasons. From the theoretical perspective, it goes distribution. We study mean estimation and beyond the typical i.i.d. setting and raises many inter- linear regression under general conditions, esting open questions, even on basic topics like mean and analyze a simple and computationally ef- estimation for Gaussians. It can also be viewed as ficient method based on iteratively trimming a generalization of the traditional mixture modeling, samples and re-estimating the parameter on since the number of distinct mixture components can the trimmed sample set. We show that the grow with the number of samples. From the practical method in logarithmic iterations outputs an perspective, many modern applications have various estimation whose error only depends on the forms of heterogeneity, for which the i.i.d. assumption noise level of the dαne-th noisiest data point can lead to bad modeling of their data. The entan- where α is a constant and n is the sample gled single-sample setting provides potentially better size. This means it can tolerate a constant modeling. This is particularly the case for applications fraction of high-noise points. These are the where we have no control over the noise levels of the first such results under our general conditions samples. For example, the images taken by self-driving with computationally efficient estimators. It cars can have varying degrees of noise due to chang- also justifies the wide application and empir- ing weather or lighting conditions. Similarly, signals ical success of iterative trimming in practice. collected from sensors on the Internet of Things can Our theoretical results are complemented by come with interferences from a changing environment. experiments on synthetic data. Though theoretically interesting and practically im- portant, few studies exist in this setting. Chierichetti et al. (2014) considered the mean estimation for en- 1 INTRODUCTION tangled Gaussians and showed the existence of a gap between estimation error rates of the best possible es- This work considers the novel parameter estimation timator in this setting and the maximum likelihood setting called entangled single-sample distributions. estimator when the variances are known. Pensia et al. Different from the typical i.i.d. setting, here we have (2019) considered means estimation for symmetric, n data points that are independent, but each is from unimodal distributions including the symmetric mul- a different distribution. These distributions are en- tivariate case (i.e., the distributions are radially sym- tangled in the sense that they share some common metric) with sharpened bounds, and provided exten- parameter, and our goal is to estimate the common sive discussion on the performance of their estimators parameter. For example, in the problem of mean es- in different configurations of the variances. These ex- timation for entangled single-sample distributions, we isting results focus on specific family of distributions or rd focus on the case where most samples are \high-noise" Proceedings of the 23 International Conference on Artifi- points. cial Intelligence and Statistics (AISTATS) 2020, Palermo, Italy. PMLR: Volume 108. Copyright 2020 by the au- On the contrary, we focus on the case with a constant thor(s). fraction of "high-noise" points, which is more inter- Learning Entangled Single-Sample Distributions via Iterative Trimming esting in practice. We study multivariate mean esti- k-shortest gap, k-median estimators) and hybrid esti- mation and linear regression under more general con- mator, which combines Median estimator with Short- ditions and analyze a simple and efficient estimator est Gap or Modal Interval estimator. They also dis- based on iterative trimming. The iterative trimming cussed slight relaxation of the symmetry assumption idea is simple: the algorithm keeps an iterate and re- and provided extensions to linear regression. Our work peatedly refines it; each time it trims a fraction of bad considers mean estimation and linear regression under points based on the current iterate and then uses the more general conditions and analyzes a simpler estima- trimmed sample set to compute the next iterate. It tor. However, our results are not directly comparable is computationally very efficient and widely used in to the existing ones above, since those focus on the practice as a heuristic for handling noisy data. It can case where most of the points have high noise or have also be viewed as an alternating-update version of the extra constraints on distributions are assumed. For classic trimmed estimator (e.g., Huber (2011)) which the constrained distributions, our results are weaker typically takes exponential time: than the existing ones. See the detailed discussion in the remarks after our theorems. ^ X θ = arg min Lossi(θ) θ2Θ;S⊆[n];jSj=dαne This setting is also closely related to robust estimation, i2S which have been extensively studied in the literature where Θ is the feasible set for the parameter θ to be of both classic statistics and machine learning theory. estimated, dαne is the size of the trimmed sample set Robust mean estimation. There are several classes S, and Lossi(θ) is the loss of θ on the i-th data point of data distribution models for robust mean estima- (e.g., `2 error for linear regression). tors. The most commonly addressed is adversarial For mean estimation, only assuming the distributions contamination model, whose origin can be traced back have a common mean and bounded covariances, we to the malicious noise model by Valiant (1985) and the show that the iterative trimming method in logarith- contamination model by Huber (2011). Under contam- mic iterations outputs a solution whose error only de- ination, mean estimation has been investigated in Di- pends on the noise level of the dαne-th noisiest point akonikolas et al. (2017, 2019a); Cheng et al. (2019). for α ≥ 4=5. More precisely, the error only depends on Another related model is the mixture of distributions. the dαne-th largest value among all the norms of the There has been steady progress in algorithms for lean- n covariance matrices. This means the method can ing mixtures, in particular, leaning Gaussian mixtures. tolerate a 1=5 fraction of \high-noise" points. We also Starting from Dasgupta (1999), a rich collection of re- provide a similar result for linear regression, under a sults are provided in many studies, such as Sanjeev regularity condition that the explanatory variables are and Kannan (2001); Achlioptas and McSherry (2005); sufficiently spread out in different directions (satisfied Kannan et al. (2005); Belkin and Sinha (2010a,b); by typical distributions like Gaussians). As far as we Kalai et al. (2010); Moitra and Valiant (2010); Di- know, these are the first such results of iterative trim- akonikolas et al. (2018a). ming under our general conditions in the entangled Robust regression. Robust Least Squares Regres- single-sample distributions setting. These results also sion (RLSR) addresses the problem of learning regres- theoretically justify the wide application and empir- sion coefficients in the presence of corruptions in the ical success of the simple iterative trimming method response vector. A class of robust regression estimator in practice. Experiments on synthetic data provide solving RLSR is Least Trimmed Square (LTS) estima- positive support for our analysis. tor, which is first introduced by Rousseeuw (1984) and has high breakdown point. The algorithm solutions of 2 RELATED WORK LTS are investigated in H¨ossjer(1995); Rousseeuw and Van Driessen (2006); Shen et al. (2013) for the linear Entangled distributions. This setting is first stud- regression setting. Recently, for robust linear regres- ied by Chierichetti et al. (2014), which considered sion in the adversarial setting (i.e., a small fraction mean estimation for entangled Gaussians and pre- of responses are replaced by adversarial values), there sented a algorithm combining the k-median and the is a line of work providing algorithms with theoretical k-shortest gap algorithms. It also showed the exis- guarantees following the idea of LTS, e.g., Bhatia et al. tence of a gap between the error rates of the best pos- (2015); Vainsencher et al. (2017); Yang et al. (2018) for sible estimator in this setting and the maximum likeli- example. For robust linear regression in the adversary hood estimator when the variances are known. Pensia setting where both explanatory and response variables et al. (2019) considered a more general class of distri- can be replaced by adversarial values, a line of work butions (unimodal and symmetric) and provided anal- provided algorithms and guarantees, e.g., Diakoniko- ysis on both individual estimator (r-modal interval, las et al. (2018b); Prasad et al. (2018); Klivans et al. Hui Yuan, Yingyu Liang (2018); Shen and Sanghavi (2019), while some others Algorithm 1 Iterative Trimmed MEAN (ITM) like Chen et al. (2013); Balakrishnan et al. (2017); Liu n Input: Samples fxigi=1 ; number of rounds T; frac- et al. (2018) considered the high-dimensional scenario. tion of samples α 1 Pn 1: µ0 n i=1 xi 3 MEAN ESTIMATION 2: for t = 0; ··· ;T − 1 do 3: Choose samples with smallest current loss: d Suppose we have n independent samples xi ∼ Fi 2 R , ? X 2 d 2 N , where the mean vector and the covariance ma- St arg min kxi − µtk2 S:jSj=dαne trix of each distribution Fi exist.
Load more

Recommended publications
  • Efficient Inflation Estimation
    1 Efficient Inflation Estimclticn by Michael F. Bryan, Stephen G. Cecchetti, and Rodney L. Wiggins II Working Paper 9707 EFFICIENT JIVFLATION ESTIMATION by Michael F. Bryan, Stephen G. Cecchetti, and Rodney L. Wiggins I1 Michael F. Bryan is assistant vice president and economist at the Federal Reserve Bank of Cleveland. Stephen G. Cecchetti is executive vice president and director of research at the Federal Reserve Bank of New York and a research associate at the National Bureau of Econonlic Research. Rodney L. Wiggins I1 is a member of the Research Department of the Federal Reserve Bank of Cleveland. The authors gratefully acknowledge the comments and assistance of Todd Clark, Ben Craig, and Scott Roger. Working papers of the Federal Reserve Bank of Cleveland are preliminary materials circulated to stimulate discussion and critical comment. The views stated herein are those of the authors and are not necessarily those of the Federal Reserve Bank of Cleveland or of the Board of Governors of the Federal Reserve System. Federal Reserve Bank of Cleveland working papers are distributed for the purpose of promoting discussion of research in progress. These papers may not have been subject to the formal editorial review accorded official Federal Reserve Bank of Cleveland publications. Working papers are now available electronically through the Cleveland Fed's home page on the World Wide Web: http:Ilwww.clev.frb.org. August 1997 Abstract This paper investigates the use of trimmed means as high-frequency estimators of inflation. The known characteristics of price change distributions, specifically the . observation that they generally exhibit high levels of kurtosis, imply that simple averages of price data are unlikely to produce efficient estimates of inflation.
    [Show full text]
  • A Globally Optimal Solution to 3D ICP Point-Set Registration
    1 Go-ICP: A Globally Optimal Solution to 3D ICP Point-Set Registration Jiaolong Yang, Hongdong Li, Dylan Campbell, and Yunde Jia Abstract—The Iterative Closest Point (ICP) algorithm is one of the most widely used methods for point-set registration. However, being based on local iterative optimization, ICP is known to be susceptible to local minima. Its performance critically relies on the quality of the initialization and only local optimality is guaranteed. This paper presents the first globally optimal algorithm, named Go-ICP, for Euclidean (rigid) registration of two 3D point-sets under the L2 error metric defined in ICP. The Go-ICP method is based on a branch-and-bound (BnB) scheme that searches the entire 3D motion space SE(3). By exploiting the special structure of SE(3) geometry, we derive novel upper and lower bounds for the registration error function. Local ICP is integrated into the BnB scheme, which speeds up the new method while guaranteeing global optimality. We also discuss extensions, addressing the issue of outlier robustness. The evaluation demonstrates that the proposed method is able to produce reliable registration results regardless of the initialization. Go-ICP can be applied in scenarios where an optimal solution is desirable or where a good initialization is not always available. Index Terms—3D point-set registration, global optimization, branch-and-bound, SE(3) space search, iterative closest point F 1 INTRODUCTION given an initial transformation (rotation and transla- Point-set registration is a fundamental problem in tion), it alternates between building closest-point cor- computer and robot vision.
    [Show full text]
  • Trimmed Mean Group Estimation
    Trimmed Mean Group Estimation Yoonseok Lee∗ Donggyu Sul† Syracuse University University of Texas at Dallas May 2020 Abstract This paper develops robust panel estimation in the form of trimmed mean group estimation. It trims outlying individuals of which the sample variances of regressors are either extremely small or large. As long as the regressors are statistically independent of the regression coefficients, the limiting distribution of the trimmed estimator can be obtained in a similar way to the standard mean group estimator. We consider two trimming methods. The first one is based on the order statistic of the sample variance of each regressor. The second one is based on the Mahalanobis depth of the sample variances of regressors. We apply them to the mean group estimation of the two-way fixed effect model with potentially heterogeneous slope parameters and to the commonly correlated regression, and we derive limiting distribution of each estimator. As an empirical illustration, we consider the effect of police on property crime rates using the U.S. state-level panel data. Keywords: Trimmed mean group estimator, Robust estimator, Heterogeneous panel, Two- way fixed effect, Commonly correlated estimator JEL Classifications: C23, C33 ∗Address: Department of Economics and Center for Policy Research, Syracuse University, 426 Eggers Hall, Syracuse, NY 13244. E-mail: [email protected] †Address: Department of Economics, University of Texas at Dallas, 800 W. Campbell Road, Richardson, TX 75080. E-mail: [email protected] 1Introduction Though it is popular in practice to impose homogeneity in panel data regression, the homogeneity restriction is often rejected (e.g., Baltagi, Bresson, and Pirotte, 2008).
    [Show full text]
  • Programs Tramo (Time Series Regression with Arima Noise
    PROGRAMS TRAMO (TIME SERIES REGRESSION WITH ARIMA NOISE, MISSING OBSERVATIONS AND OUTLIERS) AND SEATS (SIGNALEXTRACTIONS AREMA TIME SERIES) INSTRUCTIONS FOR THE USER (BETA VERSION: JUNIO 1997) Víctor Gómez * Agustín Maravall ** SGAPE-97001 Julio 1997 * Ministerio de Economía y Hacienda ** Banco de España This software is available at the following Internet address: http://www.bde.es The Working Papers of the Dirección General de Análisis y Programación Presupuestaria are not official statements of the Ministerio de Economía y Hacienda Abstract Brief summaries and user Instructions are presented for the programs-* TRAMO ("Time 'Series Regression with ARTMA Noise, Missing Observations and Outliers") and SEATS ("Signal Extraction in ARIMA Time Series"). TRAMO is a program for estimation and forecasting of regression models with possibly nonstationary (ARIMA) errors and any sequence of missing val- ues. The program interpolates these values, identifies and corrects for several types of outliers, and estimates special effects such as Trading Day and Easter and, in general, intervention variable type of effects. Fully automatic model identification and outlier correction procedures are available. SEATS is a program for estimation of unobserved components in time se- ries following the so-called ARlMA-model-based method. The trend, seasonal, irregular, and cyclical components are estimated and forecasted with signal extraction techniques applied to ARIMA models. The standard errors of the es- timates and forecasts are obtained and the model-based structure is exploited to answer questions of interest in short-term analysis of the data. The two programs are structured so as to be used together both for in- depth analysis of a few series (as presently done at the Bank of Spain) or for automatic routine applications to a large number of series (as presently done at Eurostat).
    [Show full text]
  • Power-Shrinkage and Trimming: Two Ways to Mitigate Excessive Weights
    ASA Section on Survey Research Methods Power-Shrinkage and Trimming: Two Ways to Mitigate Excessive Weights Chihnan Chen1, Nanhua Duan2, Xiao-Li Meng3, Margarita Alegria4 Boston University1 UCLA2 Harvard University3 Cambridge Health Alliance and Harvard Medical School4 Abstract 2 Power-Shrinkage and Trimming Large-scale surveys often produce raw weights with very The survey weights are constructed as the reciprocals of large variations. A standard approach is to perform the selection probabilities. Let w be the survey weight, n some form of trimming, as a way to reduce potentially be the sample size, and y be the variable of interest. Let large variances in various survey estimators. The amount p 2 [0; 1] be the power-shrinkage parameter and T 2 [0; 1] of trimming is usually determined by considerations of be the trimming threshold de¯ned in terms of percentile. bias-variance trade-o®. While bias-variance trade-o® is The original weighted estimator, power-shrinkage estima- a sound principle, the trimming method itself is popular tor and the trimmed estimator for the expectation, E(y), only because of its simplicity, not because it has statisti- are given by cally desirable properties. In this paper we investigate a Xn more principled method by shrinking the variability of the wiyi log of the weights. We use the same mean squared error ² Original weighted estimator :y ¹ = i=1 (MSE) criterion to determine the amount of shrinking. w Xn The shrinking on the log scale implies shrinking weights wi by a power parameter p between [0,1]. Our investiga- i=1 tion suggests an empirical way to predict the optimal Xn choice.
    [Show full text]
  • 26 Sep 2019 the Trimmed Mean in Non-Parametric Regression
    The Trimmed Mean in Non-parametric Regression Function Estimation Subhra Sankar Dhar1, Prashant Jha2, and Prabrisha Rakhshit3 1Department of Mathematics and Statistics, IIT Kanpur, India 2Department of Mathematics and Statistics, IIT Kanpur, India 3Department of Statistics, Rutgers University, USA September 27, 2019 Abstract This article studies a trimmed version of the Nadaraya-Watson estimator to estimate the unknown non-parametric regression function. The characterization of the estimator through minimization problem is established, and its pointwise asymptotic distribution is derived. The robustness property of the proposed estimator is also studied through breakdown point. Moreover, as the trimmed mean in the location model, here also for a wide range of trim- ming proportion, the proposed estimator poses good efficiency and high breakdown point for various cases, which is out of the ordinary property for any estimator. Furthermore, the usefulness of the proposed estimator is shown for three benchmark real data and various simulated data. Keywords: Heavy-tailed distribution; Kernel density estimator; L-estimator; Nadaraya-Watson estimator; Robust estimator. arXiv:1909.10734v2 [math.ST] 26 Sep 2019 1 Introduction We have a random sample (X1,Y1),..., (Xn,Yn), which are i.i.d. copies of (X,Y ), and the regres- sion of Y on X is defined as Yi = g(Xi)+ ei, i =1, . , n, (1) where g( ) is unknown, and e are independent copies of error random variable e with E(e X)=0 · i | and var(e X = x)= σ2(x) < for all x. Note that the condition E(e X) = 0 is essentially the | ∞ | 2000 AMS Subject Classification: 62G08 1 identifiable condition for mean regression, and it varies over the different procedures of regression.
    [Show full text]
  • Measuring Core Inflation in India: an Application of Asymmetric-Trimmed Mean Approach
    Measuring Core Inflation in India: An Application of Asymmetric-Trimmed Mean Approach Motilal Bicchal* Naresh Kumar Sharma ** Abstract The paper seek to obtain an optimal asymmetric trimmed means-core inflation measure in the class of trimmed means measures when the distribution of price changes is leptokurtic and skewed to the right for any given period. Several estimators based on asymmetric trimmed mean approach are constructed, and evaluated by the conditions set out in Marques et al. (2000). The data used in the study is the monthly 69 individual price indices which are constituent components of Wholesale Price Index (WPI) and covers the period, April 1994 to April 2009, with 1993-94 as the base year. Results of the study indicate that an optimally trimmed estimator is found when we trim 29.5 per cent from the left-hand tail and 20.5 per cent from the right-hand tail of the distribution of price changes. Key words: Core inflation, underlying inflation, asymmetric trimmed mean JEL Classification: C43, E31, E52 * Ph.D Research scholar and ** Professor at Department of Economics, University of Hyderabad, P.O Central University, Hyderabad-500046, India. Corresponding author: [email protected] We are very grateful to Carlos Robalo Marques of Banco de Portugal for providing the program for computing the asymmetric trimmed-mean as well as for his valuable comments and suggestions. We also wish to thank Bandi Kamaiah for going through previous draft and for offering comments, which improved the presentation of this paper. 1. Introduction Various approaches to measuring core inflation have been discussed in the literature.
    [Show full text]
  • Robust Statistics
    Robust statistics Stéphane Paltani Why robust statistics? Removal of data Robust statistics L-estimators and some non-parametric statistics R-estimators M-estimators Stéphane Paltani ISDC Data Center for Astrophysics Astronomical Observatory of the University of Geneva Statistics Course for Astrophysicists, 2010–2011 Outline Robust statistics Stéphane Paltani Why robust statistics? Removal of data Why robust statistics? L-estimators R-estimators Removal of data M-estimators L-estimators R-estimators M-estimators Caveat Robust statistics Stéphane Paltani Why robust statistics? Removal of data The presentation aims at being user-focused and at L-estimators presenting usable recipes R-estimators Do not expect a fully mathematically rigorous description! M-estimators This has been prepared in the hope to be useful even in the stand-alone version Please provide me feedback with any misconception or error you may find, and with any topic not addressed here but which should be included Outline Robust statistics Stéphane Paltani Why robust statistics? Why robust statistics? First example General concepts First example Data representation General concepts Removal of data Data representation L-estimators R-estimators M-estimators Removal of data L-estimators R-estimators M-estimators Example: The average weight of opossums Robust statistics Stéphane Paltani Why robust statistics? First example General concepts Data representation Removal of data L-estimators R-estimators M-estimators I Find a group of N opossums I Weight them: wi ; i = 1;:::; N
    [Show full text]
  • Configurable FPGA-Based Outlier Detection for Time Series Data
    Configurable FPGA-Based Outlier Detection for Time Series Data by Ghazaleh Vazhbakht, B.Sc. A thesis submitted to the Faculty of Graduate and Postdoctoral Affairs in partial fulfillment of the requirements for the degree of Master of Applied Science in Electrical and Computer Engineering Ottawa-Carleton Institute for Electrical and Computer Engineering Carleton University Ottawa, Ontario c 2017 Ghazaleh Vazhbakht TABLE OF CONTENTS Page LIST OF FIGURES iv LIST OF TABLES v LIST OF ACRONYMS vi ACKNOWLEDGMENTS vii ABSTRACT viii 1 Introduction 1 1.1 Outlier Detection in Time Series . 2 1.2 Thesis Objective and Contributions . 3 2 Background 5 2.1 Definition of Outlier . 5 2.2 Outlier Detection . 6 2.3 Scope of Work . 8 2.3.1 Types of Outliers . 10 2.3.2 Techniques Based on Outlier Aware Prediction Models . 13 2.3.3 Managing Detected Outliers . 14 2.4 FPGA-based Outlier Detection Techniques . 15 3 Theory and Reduction to Practice 20 3.1 ARMA Modeling . 21 3.1.1 Model Selection . 23 3.1.2 Parameter Estimation . 26 3.1.3 Model Checking . 35 3.2 Test Statistics . 35 3.2.1 Critical Value . 36 3.3 Outlier Detection Algorithm . 37 3.3.1 Estimating Outliers Effects . 39 3.3.2 Locating and Identifying Outliers . 41 3.3.3 Residual Standard Deviation . 42 ii 3.3.4 The Procedure of Joint Estimation of Model Parameters and Outlier Effects . 43 4 Design and Implementation 48 4.1 Profiling the Outlier Detection Algorithm . 49 4.2 MATLAB Implementation of the Simplified Algorithm . 50 4.3 Hardware Implementation of the Outlier Detection Algorithm .
    [Show full text]
  • Outlier Detection and Trimmed Estimation for General Functional Data
    Outlier detection and trimmed estimation for general functional data Daniel Gervini Department of Mathematical Sciences University of Wisconsin–Milwaukee P.O. Box 413, Milwaukee, WI 53201 [email protected] December 3, 2012 arXiv:1001.1014v2 [stat.ME] 30 Nov 2012 Abstract This article introduces trimmed estimators for the mean and covariance function of general functional data. The estimators are based on a new measure of “outlying- ness” or data depth that is well defined on any metric space, although this paper focuses on Euclidean spaces. We compute the breakdown point of the estimators and show that the optimal breakdown point is attainable for the appropriate choice of tuning parameters. The small-sample behavior of the estimators is studied by simulation, and we show that they have better outlier-resistance properties than alternative estimators. This is confirmed by two real-data applications, that also show that the outlyingness measure can be used as a graphical outlier-detection tool in functional spaces where visual screening of the data is difficult. Key Words: Breakdown Point; Data Depth; Robust Statistics; Stochastic Pro- cesses. 1 Introduction Many statistical applications today involve data that does not fit into the classi- cal univariate or multivariate frameworks; for example, growth curves, spectral curves, and time-dependent gene expression profiles. These are samples of func- tions, rather than numbers or vectors. We can think of them as realizations of a stochastic process with sample paths in L 2(R), the space of square-integrable functions. The statistical analysis of function-valued data has received a lot of at- tention in recent years (see e.g.
    [Show full text]
  • Robust Mean Estimation for Real-Time Blanking in Radioastronomy
    ROBUST MEAN ESTIMATION FOR REAL-TIME BLANKING IN RADIOASTRONOMY Philippe Ravier1, Cedric Dumez-Viou1−2 1Laboratory of Electronics, Signals, Images, Polytech'Orleans´ - University of Orleans 12, rue de Blois, BP 6744 Cedex 2, F-45067 Orleans,´ France 2Observatoire de Paris/(CNRS No.704), Station de radioastronomie, F-18330 Nanc¸ay, France phone: + (33).2.38.49.48.63, fax: + (33).2.38.41.72.45, email: [email protected] web: www.obs-nancay.fr/rfi ABSTRACT assumption does not hold anymore and the exact PDF must be taken into account for precise threshold derivation. Radio astronomy observations suffer from strong interfer- ences that need to be blanked both in the time and frequency domains. In order to achieve real-time computations, in- 2. PROCEDURE USED FOR REAL-TIME terference detection is made by simple thresholding. The In order to achieve real-time computation, a simple thresh- threshold value is linked to the mean estimation of the power olding drives the time and/or frequency blanking decision. 2 of clean observed data that follows a χ distribution. Spu- The problem is formulated as the following binary hy- rious values insensitivity is obtained by replacing mean by pothesis test: robust mean. A theoretical study of the variance of three es- timators of the mean is presented. The study leads to a prac- tical trimmed percentage that is chosen for the robust mean. H0 : the SOI only is present, H1 : the SOI is polluted with additive interference. 1. INTRODUCTION The SOI is supposed to be Gaussian. On the other hand, Radioastronomy benefits from protected bands for the study no a priori information about the interference PDF is avail- of radio-sources.
    [Show full text]
  • To Trim Or Not to Trim?
    To trim or not to trim? An application of a trimmed mean inflation estimator to the United Kingdom Hasan Bakhshi and Tony Yates We are very grateful to Andrew Brigden for the work he did developing computer programs used in this paper. Comments from Bill Allen, Michael Bryan, Steve Cecchetti, Mark Cornelius, Alec Chrystal, Alastair Cunningham, Mervyn King, Danny Quah and Scott Roger materially improved this paper. All remaining errors are ours alone. The views expressed in this paper are those of the authors and not necessarily those of the Bank of England. Issued by the Bank of England, London, EC2R 8AH to which requests for individual copies should be addressed: envelopes should be marked for the attention of the Publications Group (Telephone 0171- 601 4030). Working papers are also available on the Bank’s Internet site at http://www.bankofengland.co.uk. Bank of England 1999 ISSN number 1368-5562 2 Contents Abstract 5 1 Introduction 7 2 The theory of the trimmed mean 8 (i) The case in favour 8 (ii) The case against 10 3 Statistical theory 13 4 Application to the United Kingdom 15 (i) The optimal trim for the United Kingdom 15 (ii) Robustness of the trimmed mean calculation 19 (iii) Problems with Bryan et al’s methodology 28 5 Conclusions 33 References 36 3 Abstract Although the target of monetary policy is clear, there have been suggestions that the conduct of monetary policy is improved by monitoring ‘trimmed mean’ inflation rates, the mean of some central portion of the distribution of price changes. This paper assesses critically the theoretical and empirical arguments for trimming, and applies Bryan et al’s (1997) concept of the ‘optimal trim’ to the United Kingdom.
    [Show full text]