DOCSLIB.ORG

Hellinger Distance Based Drift Detection for Nonstationary Environments

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Hellinger Distance Based Drift Detection for Nonstationary Environments Hellinger Distance Based Drift Detection for Nonstationary Environments Gregory Ditzler and Robi Polikar Dept. of Electrical & Computer Engineering Rowan University Glassboro, NJ, USA [email protected], [email protected] Abstract—Most machine learning algorithms, including many sion boundaries as a concept change, whereas gradual changes online learners, assume that the data distribution to be learned is in the data distribution as a concept drift. However, when the fixed. There are many real-world problems where the distribu- context does not require us to distinguish between the two, we tion of the data changes as a function of time. Changes in use the term concept drift to encompass both scenarios, as it is nonstationary data distributions can significantly reduce the ge- usually the more difficult one to detect. neralization ability of the learning algorithm on new or field data, if the algorithm is not equipped to track such changes. When the Learning from drifting environments is usually associated stationary data distribution assumption does not hold, the learner with a stream of incoming data, either one instance or one must take appropriate actions to ensure that the new/relevant in- batch at a time. There are two types of approaches for drift de- formation is learned. On the other hand, data distributions do tection in such streaming data: in passive drift detection, the not necessarily change continuously, necessitating the ability to learner assumes – every time new data become available – that monitor the distribution and detect when a significant change in some drift may have occurred, and updates the classifier ac- distribution has occurred. In this work, we propose and analyze a cording to the current data distribution, regardless whether feature based drift detection method using the Hellinger distance drift actually did occur. In active drift detection, the algorithm to detect gradual or abrupt changes in the distribution. continuously and explicitly monitors the data to detect if and Keywords-concept drift; nonstationary environments; drift when drift occurs. If – and only if – the drift is detected, the detection algorithm takes an appropriate action, such as updating the classifier with the most recent data or simply creating a new I. INTRODUCTION classifier to learn the current data. The drift detection method Detecting change in data distributions is an important prob- presented in this work is appropriate for an active drift detec- lem in data mining and machine learning algorithms due to tion framework. increasing number of applications that are governed by such This paper is organized as follows: Section II provides an data. Any algorithm that does not make the necessary adjust- overview of drift detection algorithms, followed by our moti- ments to changes in data distribution will necessarily fail to vation for using Hellinger distance as a metric, as well as a provide satisfactory generalization on future data, if such data detailed explanation of the approach for the proposed algo- do not come from the distribution on which the algorithm was rithm in Section III. Section IV presents results obtained by originally trained. For example, consider an application that the proposed approach on several real world & synthetic data- tracks a user’s web browsing habits to determine which ads sets. Finally, Section V provides concluding remarks and are most relevant for that user’s interest. User interests are future work. known to change – or drift – over time. In such cases, certain II. BACKGROUND ads in which the customer used to express interest may no longer be relevant. Thus, an algorithm designed to determine Concept drift algorithms are usually associated with incre- the relevant ads must be able to monitor the customer’s brows- mental learning of streaming data, where new datasets become ing habits and determine when there is change in the available in batches or in an instance-by-instance basis, result- customer’s interest. Applications that call for an effective ing in batch or online learning, respectively [1]. Given such change or drift detection algorithm can be expanded: analysis data, the (active) drift detection itself can be parametric or of electricity demands or pricing, financial or climate data are non-parametric, depending on whether a specific underlying all examples of applications with nonstationary data distribu- distribution is assumed. Many parametric algorithms use a tions, where change or drift detection is needed so that the CUSUM (cumulative sum) based mechanism, which is tradi- learner can take an appropriate action. tionally used for control charts in detecting nonstationary For the purposes of this paper, we define a sudden or abrupt changes in process data [2-4]. A series of successful imple- change in the underlying data distribution that alters the deci- mentations of this approach are proposed by Alippi & Roveri, including CI-CUSUM [5;6], a pdf free extension of the tradi- This material is based on work supported by the National Science Founda- tion (NSF) under Grant No: ECCS-0926159. tional CUSUM. Recently, Alippi et al. also introduced the 978-1-4244-9931-1/11/$26.00 ©2011 IEEE intersection of confidence intervals (ICI) rule for drift detec- change between the distributions of data at two subsequent tion [7]. Some drift detection approaches such as the Early time stamps. Drift Detection Method (EDDM) [8] and similar approaches In contrast to EDDM and other similar approaches that rely [9], do not make any assumptions on feature distribution, but on classifier error [8;9], the proposed Hellinger distance drift rather monitor a classifiers’ accuracy or some distance metric detection method (HDDDM) is a feature based drift detection to detect drift. Cielslak & Chawla suggest Hellinger distance, method, using the Hellinger distance between current data dis- not for detecting concept drift in an incremental learning set- tribution and a reference distribution that is updated as new ting, but rather to detect bias between training and test data data are received. The Hellinger distance is an example of distributions [10]. Using a non-parametric statistical test, the ͚divergence measure, similar to the Kullback-Leibler (KL) authors measure the significance between the probability esti- divergence. However, unlike the KL-divergence the Hellinger mates of the classifier on a validation set (carved out from divergence is a symmetric metric. Furthermore, unlike most training data) and the corresponding test dataset [10]. To do other distance metrics, the Hellinger distance is a bounded dis- so, a baseline comparison is first made by calculating the Hel- tance measure: for two distributions with probability mass linger distance between the original training and test datasets. functions (or histograms representing these distributions) ͊ Bias is then injected into the testing set. A baseline Hellinger and ͋Q the Hellinger distance is ʚ͊Q ͋ʛ ǨƫRQ:TƯ . If distance (i.e. when no bias is present between training/testing ʚ͊Q ͋ʛ ƔR, the two probability mass functions are com- sets) and the distance after bias is injected into the dataset are pletely overlapping and hence identical. If ʚ͊Q ͋ʛ Ɣ :T, the observed. two probability mass functions are completely divergent (i.e. We extend this approach to concept drift in an incremental there is no overlap). learning setting, where new datasets are presented in batches As an example, consider a two-class rotating mixture of over time. We do not use the original dataset as the baseline Gaussians with class centers moving in a circular pattern (Fig. distribution against which other data distributions are com- 1), with each distribution in the mixture corresponding to a pared, nor do we inject any bias into future distributions, as different class label. The class means can be given by the pa- done in [10], but rather we monitor the magnitude of the ʚ ʛ ʚ ʛ ʚ ʛ / Ɣ ʞ!-1 Q1', ʟ / ƔƎ/ change in Hellinger distance between a new distribution and a rametric equations ͥ / / , ͦ ͥ , Ɣ ͦ_ ͨ baseline distribution, which is updated every time drift is de- / , with fixed class covariance matrices given as tected. 9ͥ Ɣ9ͦ ƔRTWdz̏, where ͗ is the number of cycles, ͨ is the (integer valued) time stamp that iterates from zero to ͈ƎS, III. APPROACH and ̏ is a 2x2 identity matrix. Fig. 2 shows the evolution of the We begin this section by introducing the motivation for us- Hellinger distance computed between the datasets (Ď&) gener- ing the Hellinger distance as a measure that can be applied to ated with respect to ͥ and & where ͟ƔTQUQ͈ƎS. The drift detection in an incremental learning setting. Next, we Hellinger distance is capable of displaying the relevance or the present the proposed Hellinger Distance Drift Detection Me- closeness of a new dataset (Ď&) to a baseline dataset (Ďͥ) as thod (HDDDM). According to four criteria suggested by shown in Fig. 2. We plot the Hellinger distance for the diver- Kuncheva [3], HDDDM can be categorized as follows: gence of the datasets for class !ͥ, class !ͦ, and the entire data separately. The Hellinger distance varies as begins to Data chunks: batch based evolve. We observe that when ͥ and & are the same (or very Information used: raw features (not based on classifier similar) the Hellinger distance is small as observed at ͨƔ performance) ʚ/ʛ ʚͥʛ ʚ/ʛ ʜRQ TRRQ VRRQ XRRʝ . This is when ͥ Ɣͥ and ͦ Ɣ Change detection mode: explicit ʚ/ʛ ʚͥʛ Ǝ ƔƎ . We observe that the Hellinger distance com- Classifier-specific vs. classifier-free: classifier free ͥ ͥ puted for all data (entire dataset) repeats every 100 time The algorithm uses a hypothesis testing-like statistical ap- stamps, twice as often compared to class specific distributions, proach to determine if there is enough evidence to suggest that t=25 the data instances are being drawn from different distributions.
Load more

Recommended publications
  • Hellinger Distance-Based Similarity Measures for Recommender Systems
    Hellinger Distance-based Similarity Measures for Recommender Systems Roma Goussakov One year master thesis Ume˚aUniversity Abstract Recommender systems are used in online sales and e-commerce for recommend- ing potential items/products for customers to buy based on their previous buy- ing preferences and related behaviours. Collaborative filtering is a popular computational technique that has been used worldwide for such personalized recommendations. Among two forms of collaborative filtering, neighbourhood and model-based, the neighbourhood-based collaborative filtering is more pop- ular yet relatively simple. It relies on the concept that a certain item might be of interest to a given customer (active user) if, either he appreciated sim- ilar items in the buying space, or if the item is appreciated by similar users (neighbours). To implement this concept different kinds of similarity measures are used. This thesis is set to compare different user-based similarity measures along with defining meaningful measures based on Hellinger distance that is a metric in the space of probability distributions. Data from a popular database MovieLens will be used to show the effectiveness of different Hellinger distance- based measures compared to other popular measures such as Pearson correlation (PC), cosine similarity, constrained PC and JMSD. The performance of differ- ent similarity measures will then be evaluated with the help of mean absolute error, root mean squared error and F-score. From the results, no evidence were found to claim that Hellinger distance-based measures performed better than more popular similarity measures for the given dataset. Abstrakt Titel: Hellinger distance-baserad similaritetsm˚attf¨orrekomendationsystem Rekomendationsystem ¨aroftast anv¨andainom e-handel f¨orrekomenderingar av potentiella varor/produkter som en kund kommer att vara intresserad av att k¨opabaserat p˚aderas tidigare k¨oppreferenseroch relaterat beteende.
    [Show full text]
  • On Measures of Entropy and Information
    On Measures of Entropy and Information Tech. Note 009 v0.7 http://threeplusone.com/info Gavin E. Crooks 2018-09-22 Contents 5 Csiszar´ f-divergences 12 Csiszar´ f-divergence ................ 12 0 Notes on notation and nomenclature 2 Dual f-divergence .................. 12 Symmetric f-divergences .............. 12 1 Entropy 3 K-divergence ..................... 12 Entropy ........................ 3 Fidelity ........................ 12 Joint entropy ..................... 3 Marginal entropy .................. 3 Hellinger discrimination .............. 12 Conditional entropy ................. 3 Pearson divergence ................. 14 Neyman divergence ................. 14 2 Mutual information 3 LeCam discrimination ............... 14 Mutual information ................. 3 Skewed K-divergence ................ 14 Multivariate mutual information ......... 4 Alpha-Jensen-Shannon-entropy .......... 14 Interaction information ............... 5 Conditional mutual information ......... 5 6 Chernoff divergence 14 Binding information ................ 6 Chernoff divergence ................. 14 Residual entropy .................. 6 Chernoff coefficient ................. 14 Total correlation ................... 6 Renyi´ divergence .................. 15 Lautum information ................ 6 Alpha-divergence .................. 15 Uncertainty coefficient ............... 7 Cressie-Read divergence .............. 15 Tsallis divergence .................. 15 3 Relative entropy 7 Sharma-Mittal divergence ............. 15 Relative entropy ................... 7 Cross entropy
    [Show full text]
  • Tailoring Differentially Private Bayesian Inference to Distance
    Tailoring Differentially Private Bayesian Inference to Distance Between Distributions Jiawen Liu*, Mark Bun**, Gian Pietro Farina*, and Marco Gaboardi* *University at Buffalo, SUNY. fjliu223,gianpiet,gaboardig@buffalo.edu **Princeton University. [email protected] Contents 1 Introduction 3 2 Preliminaries 5 3 Technical Problem Statement and Motivations6 4 Mechanism Proposition8 4.1 Laplace Mechanism Family.............................8 4.1.1 using `1 norm metric.............................8 4.1.2 using improved `1 norm metric.......................8 4.2 Exponential Mechanism Family...........................9 4.2.1 Standard Exponential Mechanism.....................9 4.2.2 Exponential Mechanism with Hellinger Metric and Local Sensitivity.. 10 4.2.3 Exponential Mechanism with Hellinger Metric and Smoothed Sensitivity 10 5 Privacy Analysis 12 5.1 Privacy of Laplace Mechanism Family....................... 12 5.2 Privacy of Exponential Mechanism Family..................... 12 5.2.1 expMech(;; ) −Differential Privacy..................... 12 5.2.2 expMechlocal(;; ) non-Differential Privacy................. 12 5.2.3 expMechsmoo −Differential Privacy Proof................. 12 6 Accuracy Analysis 14 6.1 Accuracy Bound for Baseline Mechanisms..................... 14 6.1.1 Accuracy Bound for Laplace Mechanism.................. 14 6.1.2 Accuracy Bound for Improved Laplace Mechanism............ 16 6.2 Accuracy Bound for expMechsmoo ......................... 16 6.3 Accuracy Comparison between expMechsmoo, lapMech and ilapMech ...... 16 7 Experimental Evaluations 18 7.1 Efficiency Evaluation................................. 18 7.2 Accuracy Evaluation................................. 18 7.2.1 Theoretical Results.............................. 18 7.2.2 Experimental Results............................ 20 1 7.3 Privacy Evaluation.................................. 22 8 Conclusion and Future Work 22 2 Abstract Bayesian inference is a statistical method which allows one to derive a posterior distribution, starting from a prior distribution and observed data.
    [Show full text]
  • An Information-Geometric Approach to Feature Extraction and Moment
    An information-geometric approach to feature extraction and moment reconstruction in dynamical systems Suddhasattwa Dasa, Dimitrios Giannakisa, Enik˝oSz´ekelyb,∗ aCourant Institute of Mathematical Sciences, New York University, New York, NY 10012, USA bSwiss Data Science Center, ETH Z¨urich and EPFL, 1015 Lausanne, Switzerland Abstract We propose a dimension reduction framework for feature extraction and moment reconstruction in dy- namical systems that operates on spaces of probability measures induced by observables of the system rather than directly in the original data space of the observables themselves as in more conventional methods. Our approach is based on the fact that orbits of a dynamical system induce probability measures over the measur- able space defined by (partial) observations of the system. We equip the space of these probability measures with a divergence, i.e., a distance between probability distributions, and use this divergence to define a kernel integral operator. The eigenfunctions of this operator create an orthonormal basis of functions that capture different timescales of the dynamical system. One of our main results shows that the evolution of the moments of the dynamics-dependent probability measures can be related to a time-averaging operator on the original dynamical system. Using this result, we show that the moments can be expanded in the eigenfunction basis, thus opening up the avenue for nonparametric forecasting of the moments. If the col- lection of probability measures is itself a manifold, we can in addition equip the statistical manifold with the Riemannian metric and use techniques from information geometry. We present applications to ergodic dynamical systems on the 2-torus and the Lorenz 63 system, and show on a real-world example that a small number of eigenvectors is sufficient to reconstruct the moments (here the first four moments) of an atmospheric time series, i.e., the realtime multivariate Madden-Julian oscillation index.
    [Show full text]
  • Issue PDF (13986
    European Mathematical Society NEWSLETTER No. 22 December 1996 Second European Congress of Mathematics 3 Report on the Second Junior Mathematical Congress 9 Obituary - Paul Erdos . 11 Report on the Council and Executive Committee Meetings . 15 Fifth Framework Programme for Research and Development . 17 Diderot Mathematics Forum . 19 Report on the Prague Mathematical Conference . 21 Preliminary report on EMS Summer School . 22 EMS Lectures . 27 European Won1en in Mathematics . 28 Euronews . 29 Problem Corner . 41 Book Reviews .....48 Produced at the Department of Mathematics, Glasgow Caledonian University Printed by Armstrong Press, Southampton, UK EDITORS Secretary Prof Roy Bradley Peter W. Michor Department of Mathematics Institut fiir Mathematik, Universitiit Wien, Strudlhof­ Glasgow Caledonian University gasse 4, A-1090 Wien, Austria. GLASGOW G4 OBA, SCOTLAND e-mail: [email protected] Editorial Team Glasgow: Treasurer R. Bradley, V. Jha, J. Gomatam, A. Lahtinen G. Kennedy, M. A. Speller, J. Wilson Department of Mathematics, P.O.Box 4 Editor - Mathematics Education FIN-00014 University of Helsinki Finland Prof. Vinicio Villani Dipartimento di Matematica e-mail: [email protected] Via Bounarroti, 2 56127 Pisa, Italy EMS Secretariat e-mail [email protected] Ms. T. Makelainen University of Helsinki (address above) Editors - Brief Reviews e-mail [email protected] I Netuka and V Soucek tel: +358-9-1912 2883 Mathematical Institute Charles University telex: 124690 Sokolovska 83 fax: +358-9-1912 3213 18600 Prague, Czech Republic e-mail: Newsletter editor [email protected] R. Bradley, Glasgow Caledonian University ( address [email protected] above) USEFUL ADDRESSES e-mail [email protected] President: Jean-Pierre Bourguignon Newsletter advertising officer IHES, Route de Chartres, F-94400 Bures-sur-Yvette, M.
    [Show full text]
  • Statistics As Both a Purely Mathematical Activity and an Applied Science NAW 5/18 Nr
    Piet Groeneboom, Jan van Mill, Aad van der Vaart Statistics as both a purely mathematical activity and an applied science NAW 5/18 nr. 1 maart 2017 55 Piet Groeneboom Jan van Mill Aad van der Vaart Delft Institute of Applied Mathematics KdV Institute for Mathematics Mathematical Institute Delft University of Technology University of Amsterdam Leiden University [email protected] [email protected] [email protected] In Memoriam Kobus Oosterhoff (1933–2015) Statistics as both a purely mathematical activity and an applied science On 27 May 2015 Kobus Oosterhoff passed away at the age of 82. Kobus was employed at contact with Hemelrijk and the encourage- the Mathematisch Centrum in Amsterdam from 1961 to 1969, at the Roman Catholic Univer- ment received from him, but he did his the- ity of Nijmegen from 1970 to 1974, and then as professor in Mathematical Statistics at the sis under the direction of Willem van Zwet Vrije Universiteit Amsterdam from 1975 until his retirement in 1996. In this obituary Piet who, one year younger than Kobus, had Groeneboom, Jan van Mill and Aad van der Vaart look back on his life and work. been a professor at Leiden University since 1965. Kobus became Willem’s first PhD stu- Kobus (officially: Jacobus) Oosterhoff was diploma’ (comparable to a masters) in dent, defending his dissertation Combina- born on 7 May 1933 in Leeuwarden, the 1963. His favorite lecturer was the topolo- tion of One-sided Test Statistics on 26 June capital of the province of Friesland in the gist J. de Groot, who seems to have deeply 1969 at Leiden University.
    [Show full text]
  • On Some Properties of Goodness of Fit Measures Based on Statistical Entropy
    IJRRAS 13 (1) ● October 2012 www.arpapress.com/Volumes/Vol13Issue1/IJRRAS_13_1_22.pdf ON SOME PROPERTIES OF GOODNESS OF FIT MEASURES BASED ON STATISTICAL ENTROPY Atif Evren1 & Elif Tuna2 1,2 Yildiz Technical University Faculty of Sciences and Literature , Department of Statistics Davutpasa, Esenler, 34210, Istanbul Turkey ABSTRACT Goodness of fit tests can be categorized under several ways. One categorization may be due to the differences between observed and expected frequencies. The other categorization may be based upon the differences between some values of distribution functions. Still the other one may be based upon the divergence of one distribution from the other. Some widely used and well known divergences like Kullback-Leibler divergence or Jeffreys divergence are based on entropy concepts. In this study, we compare some basic goodness of fit tests in terms of their statistical properties with some applications. Keywords: Measures for goodness of fit, likelihood ratio, power divergence statistic, Kullback-Leibler divergence, Jeffreys’ divergence, Hellinger distance, Bhattacharya divergence. 1. INTRODUCTION Boltzmann may be the first scientist who emphasized the probabilistic meaning of thermodynamical entropy. For him, the entropy of a physical system is a measure of disorder related to it [35]. For probability distributions, on the other hand, the general idea is that observing a random variable whose value is known is uninformative. Entropy is zero if there is unit probability at a single point, whereas if the distribution is widely dispersed over a large number of individually small probabilities, the entropy is high [10]. Statistical entropy has some conflicting explanations so that sometimes it measures two complementary conceptions like information and lack of information.
    [Show full text]
  • New Developments in Statistical Information Theory Based on Entropy and Divergence Measures
    New Developments in Statistical Information Theory Based on Entropy and Divergence Measures Edited by Leandro Pardo Printed Edition of the Special Issue Published in Entropy www.mdpi.com/journal/entropy New Developments in Statistical Information Theory Based on Entropy and Divergence Measures New Developments in Statistical Information Theory Based on Entropy and Divergence Measures Special Issue Editor Leandro Pardo MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade Special Issue Editor Leandro Pardo Universidad Complutense de Madrid Spain Editorial Office MDPI St. Alban-Anlage 66 4052 Basel, Switzerland This is a reprint of articles from the Special Issue published online in the open access journal Entropy (ISSN 1099-4300) from 2017 to 2019 (available at: https://www.mdpi.com/journal/entropy/special issues/Divergence Measures) For citation purposes, cite each article independently as indicated on the article page online and as indicated below: LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. Journal Name Year, Article Number, Page Range. ISBN 978-3-03897-936-4 (Pbk) ISBN 978-3-03897-937-1 (PDF) c 2019 by the authors. Articles in this book are Open Access and distributed under the Creative Commons Attribution (CC BY) license, which allows users to download, copy and build upon published articles, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications. The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons license CC BY-NC-ND. Contents About the Special Issue Editor ...................................... vii Leandro Pardo Reprinted from: Entropy 2019, 391, 21, doi:10.3390/e21040391 ....................
    [Show full text]
  • Evaluating Fit Using Hellinger Discrimination and Dirichlet Process Prior
    EVALUATING FIT USING HELLINGER DISCRIMINATION AND DIRICHLET PROCESS PRIOR PAPA NGOM¦, RICHARD EMILION¤ ¦ LMA - Universit¶eCheikh Anta Diop - Dakar - S¶en¶egal ¤ MAPMO - Universit¶ed'Orl¶eans- France Abstract. We evaluate the lack of ¯t in a proposed family of distributions by estimating the posterior distribution of the Hellinger distance between the true distribution generating the data and the proposed family, when the prior is a Dirichlet process. We also prove some consistency results and ¯nally we illustrate the method by using Markov Chain Monte Carlo (MCMC) techniques to compute the posterior. Our work hinges on that of K. Viele (2000-b) which holds for KL information distance. 1. Introduction Nonparametric Bayesian methods have been popular and successful in many estimation problems but their relevance in hypothesis testing situations have become of interest only recently. In particular, testing goodness-of-¯t in a Bayesian setting has received attention, for example in the works of Kass and Rafety (1995), Gelman, Meng, Stern (1996). Earlier, Beran (1977) provides a new parametric estimate procedure which is minimax robust in a small Hellinger metric neighborhood. His method is based on the proposed estimator of θ, is that b value θn which minimizes the Hellinger distance between f (fθ : θ 2 £ a speci¯ed parametric family) θbn and gbn, where gbn is a kernel density estimator. This minimum Hellinger distance estimator is used to provide a goodness-of-¯t statistic which assesses the adequacy of the parametric model. Here, we pursue this idea from a Bayesian viewpoint. Suppose we have a proposed or null family of discrete distributions D£ for an observed set of data.
    [Show full text]
  • Lecture 35: December 4 35.1 the Fundamental Statistical Distances
    36-705: Intermediate Statistics Fall 2019 Lecture 35: December 4 Lecturer: Siva Balakrishnan This and the next lecture are completely optional (and will not be on your final exam). Today we will discuss distances and metrics between distributions that are useful in statistics. I will be loose in distinguishing metrics, distances and divergences (i.e. some of these are not symmetric, and some do not satisfy triangle inequality). We will discuss them in two contexts: 1. There are metrics that are analytically useful in a variety of statistical problems, i.e. they have intimate connections with estimation and testing. 2. There are metrics that are useful in data analysis, i.e. given data we want to measure some notion of distance between (the distribution of) subsets of the data and use this in some way. There is of course overlap between these contexts and so there are distances that are useful in both, but they are motivated by slightly different considerations. 35.1 The fundamental statistical distances There are four notions of distance that have an elevated status in statistical theory. I will define them a bit crudely (trying to avoid some basic measure theory that is necessary) to try to focus on the main ideas. Throughout, we will let P; Q be two probability measures on some sample space, and we denote their densities with respect to some common underlying dominating measure be p and q. To further simplify things, I will assume that this common measure is the Lebesgue measure. 1. Total Variation: The TV distance between two distributions is: Z TV(P; Q) = sup jP (A) − Q(A)j = sup j (p(x) − q(x))dxj; A A where A is just any measurable subset of the sample space, i.e.
    [Show full text]
  • Statistical Distances and Their Role in Robustness 3
    Statistical Distances and Their Role in Robustness Marianthi Markatou, Yang Chen, Georgios Afendras and Bruce G. Lindsay Abstract Statistical distances, divergences, and similar quantities have a large his- tory and play a fundamental role in statistics, machine learning and associated sci- entific disciplines. However, within the statistical literature, this extensive role has too often been played out behind the scenes, with other aspects of the statistical problems being viewed as more central, more interesting, or more important. The behind the scenes role of statistical distances shows up in estimation, where we of- ten use estimators based on minimizing a distance, explicitly or implicitly, but rarely studying how the properties of a distance determine the properties of the estimators. Distances are also prominent in goodness-of-fit, but the usual question we ask is “how powerful is this method against a set of interesting alternatives” not “what aspect of the distance between the hypothetical model and the alternative are we measuring?” Our focus is on describing the statistical properties of some of the distance mea- sures we have found to be most important and most visible. We illustrate the robust nature of Neyman’s chi-squared and the non-robust nature of Pearson’s chi-squared statistics and discuss the concept of discretization robustness. Marianthi Markatou Department of Biostatistics, University at Buffalo, Buffalo, NY, 14214, e-mail: [email protected] Yang Chen arXiv:1612.07408v1 [math.ST] 22 Dec 2016 Department of Biostatistics, University at Buffalo, Buffalo, NY, 14214, e-mail: [email protected] Georgios Afendras Department of Biostatistics, University at Buffalo, Buffalo, NY, 14214, e-mail: [email protected] Bruce G.
    [Show full text]
  • Rare Event Analysis for Minimum Hellinger Distance Estimators Via Large Deviation Theory
    entropy Article Rare Event Analysis for Minimum Hellinger Distance Estimators via Large Deviation Theory Anand N. Vidyashankar 1,* and Jeffrey F. Collamore 2 1 Department of Statistics, George Mason University, Fairfax, VA 22030, USA 2 Department of Mathematical Sciences, University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen Ø, Denmark; [email protected] * Correspondence: [email protected] Abstract: Hellinger distance has been widely used to derive objective functions that are alternatives to maximum likelihood methods. While the asymptotic distributions of these estimators have been well investigated, the probabilities of rare events induced by them are largely unknown. In this article, we analyze these rare event probabilities using large deviation theory under a potential model misspecification, in both one and higher dimensions. We show that these probabilities decay exponentially, characterizing their decay via a “rate function” which is expressed as a convex conjugate of a limiting cumulant generating function. In the analysis of the lower bound, in particular, certain geometric considerations arise that facilitate an explicit representation, also in the case when the limiting generating function is nondifferentiable. Our analysis involves the modulus of continuity properties of the affinity, which may be of independent interest. Keywords: Hellinger distance; large deviations; divergence measures; rare event probabilities Citation: Vidyashankar, A.N.; Collamore, J.F. Rare Event Analysis 1. Introduction for Minimum Hellinger Distance In a variety of applications, the use of divergence-based inferential methods is gain- Estimators via Large Deviation ing momentum, as these methods provide robust alternatives to traditional maximum Theory. Entropy 2021, 23, 386. likelihood-based procedures. Since the work of [1,2], divergence-based methods have been https://doi.org/10.3390/e23040386 developed for various classes of statistical models.
    [Show full text]