A uniform central limit theorem for neural network based autoregressive processes with applications to change-point analysis
|
|
- Samantha Carter
- 5 years ago
- Views:
Transcription
1 A uiform cetral limit theorem for eural etwork based autoregressive processes with applicatios to chage-poit aalysis Claudia Kirch Joseph Tadjuidje Kamgaig March 6, 20 Abstract We cosider a autoregressive process with a oliear regressio fuctio that is modeled by a feedforward eural etwork. We derive a uiform cetral limit theorem which is useful i the cotext of chage-poit aalysis. We propose a test for a chage i the autoregressio fuctio which by the uiform cetral limit theorem has asymptotic power oe for a large class of alteratives icludig local alteratives. Keywords: Uiform cetral limit theorem, oparametric regressio, eural etwork, autoregressive process AMS Subject Classificatio 2000: 60F05, 62J02, 62M45 Itroductio Limit theory i geeral ca be described as the heart of probability ad mathematical statistics. Uiform cetral limit theorems for depedet radom variables i particular The work was supported by the DFG graduate college Mathematics ad ractice as well as by the DFG grat KI 443/2-. The positio of the first author was fiaced by the Stifterverbad für die Deutsche Wisseschaft by fuds of the Clausse-Simo-trust. Karlsruhe Istitute of Techology KIT, Istitute for Stochastics, Kaiserstr. 89 D 7633 Karlsruhe, Germay; claudia.kirch@kit.edu Uiversity Kaiserslauter, Departmet of Mathematics, Erwi-Schrödiger-Straße, D Kaiserslauter, Germay; tadjuidj@mathematik.ui-kl.de
2 Itroductio have give a ew dimesio to the asymptotic theory for sums of processes idexed by families of sets by givig rise to may applicatios out of reach of the classical cetral limit theorem. I this paper we cosider a oliear autoregressive time series, where we model the autoregressio fuctio by a feedforward eural etwork. Due to its uiversal approximatio property, a large class of fuctios ca be approximated by a eural etwork to ay degree of accuracy cofer e.g. White [0] or Frake et al. [4]. Therefore, this setup is very geeral ad able to model may real-life time series while at the same time beig mathematical feasible ad computatioally easier to hadle due to its parametric ature. For θ = ν 0,..., ν H, α,..., α H, β,..., β H, α j = α j,..., α jp, fx, θ = ν 0 + H ν h ψ< α h, x > +β h,. h= deotes a oe layer feedforward eural etwork with H hidde euros, <, > is the classical scalar product o R p. I this paper we assume that ψ belogs to the class of sigmoid activatio fuctios that satisfy lim ψx = 0, lim x ψx =, ψx + ψ x =..2 x A popular example is the logistic fuctio ψx = +e x which also fulfills Assumptio C.. The time series model, we have i mid i this paper, is give by X t = f X t, θ 0 + e t,.3 where X t = X t,..., X t p, θ 0 is fixed but ukow, e t idepedet of F t = σ{x u, u t } the σ-algebra geerated by the observatios up to time t. Furthermore, {e t : t } are idepedet idetically distributed radom errors with a positive variace. Stockis et al. [8] use these time series as buildig blocks i a regime-switchig model, so called CHARME-models, i the cotext of fiacial time series. I their model the duratio time i each regime is radom ad drive by a hidde Markov chai, while i classical chage-poit aalysis the duratio time is usually fixed ad determiistic. Motivated by the CHARME time series Kirch ad Tadjuidje-Kamgaig [6] developed chage-poit tests i such a setup. Chage-poit aalysis deals with the questio whether the stochastic structure of the observatios has chaged at some ukow poit i the sample, which is a importat questio i diverse areas such as ecoomy, fiace, geology, physics or quality cotrol. For a detailed discussio we refer to the book by Csörgő ad Horváth [2]. I Sectio 2 we develop a uiform cetral limit theorem ivolvig the above autoregressive time series. Based o these uiform cetral limit theorems we ca elarge i Sectio 3 the class of alteratives for which the chage-poit tests developed i Kirch ad Tadjuidje-Kamgaig [6] have asymptotic power oe allowig i particular for local chages. Fially, i Sectio 4 the proofs ca be foud. 2
3 2 Uiform Cetral Limit Theorem 2 Uiform Cetral Limit Theorem I the followig {X t } is a arbitrary statioary time series ad ot ecessarily of the form.3. The uiform cetral limit theorem for empirical processes traces back to Dudley [3] ad is fouded o otios like VC-subgraphs or coverig umbers. We will make use of this methodology. Deote by F = {fθ, ; θ Θ} the set of all feedforward etworks as defied previously with parameter θ Θ. To obtai the uiform cetral limit theorem we eed the followig assumptios. C.. The set Θ is compact. Furthermore ψ is a sigmoid activatio fuctio as i.2 which is cotiuously differetiable with bouded derivative. The latter assumptio is e.g. fulfilled for the logistic fuctio. C. 2. X t is statioary ad β-mixig with mixig coefficiet β fulfillig for some τ > 2 as k k τ/τ 2 log k 2τ /τ 2 βk 0. Furthermore E X ν < for some ν > 2. This is a classical coditio i oliear time series aalysis ad ca be derived with little effort usig the stability theory for Markov processes, see e.g. Mey ad Tweedie [7]. Ideed, this property is a cosequece of the existece of a statioary solutio that is geometric ergodic. I this situatio the assumptio o the rate is therefore fulfilled sice the time series is β-mixig with a expoetial rate. I a more geeral setup ad for the related CHARME-models this β-mixig property has bee prove by Stockis et al. [8]. For the ext assumptio we first eed to recall the defiitio of coverig umbers. Defiitio 2.. Coverig Numbers The coverig umber N ɛ, F,. is the miimal umber of balls of radius ɛ, i.e. {g : g f < ɛ}, eeded to cover F, where it is ot ecessary that f F. C. 3. Assume that for some 0 < q < 0 log N u, F,. q /2 du <. This uiform etropy coditio is classical i the cotext of weak covergece of empirical processes cf. e.g. Va der Vaart ad Weller [9]. We are ow ready to prove the uiform cetral limit theorem. 3
4 3 Chage-poit tests uder local alteratives Theorem 2.. Assume C. C.3 hold. The, { } w fx t, θ EfX t, θ; θ Θ {Gθ; θ Θ} t= i l Θ, where {Gθ} has a versio with uiformly bouded ad uiformly cotiuous paths with respect to the. 2 -orm. I particular fx t, θ EfX t, θ = O. 2. sup θ Θ t= 3 Chage-poit tests uder local alteratives We observe a time series Z t = X t {t k } + Y t, {t>k } with a possible switch from X t to Y t, at some ukow time poit k = k. The time series X t is ot ecessarily of the form.3 but we assume that it is well approximated by X t = f X t, θ 0 + e t for some θ 0 i the iterior of Θ, Y t, is a arbitrary time series. If X t follows.3, the θ 0 = θ 0. For more details o the derivatio of θ 0 we refer to Kirch ad Tadjuidje-Kamgaig [6]. The ukow parameter k is called the chage-poit ad we are iterested i the testig problem H 0 : k = vs. H : k <. Our testig procedures are based o various fuctioals of the partial sums of estimated residuals k k Ŝ k = ê t = Z t fz t, θ, 3. t=p+ t=p+ where θ is the least-squares estimator of θ 0 uder H 0, i.e. the miimizer of t=p+ Z t fz t, θ 2. Kirch ad Tadjuidje-Kamgaig [6] prove cosistecy ad asymptotic ormality of θ uder the ull hypothesis uder some mild regularity coditios - i additio to some asymptotic results uder alteratives as well as misspecificatio. If the estimator is ot i the iterior of Θ we reject the ull hypothesis as asymptotically this ca oly happe uder the alterative or if the model.3 is ot suitable for the data set at had. Cosequetly, we ca assume w.l.o.g. θ Θ, the iterior of Θ, for asymptotic power cosideratios. Typical test statistics i this cotext are give by p T, = max p<k< k p k Ŝk, T 2, q = max p<k< p q k Ŝk, 3.2 p T 3, G = max p+g<k Ŝ k Ŝk G G 4
5 3 Chage-poit tests uder local alteratives where q is a positive weight fuctios defied o 0, fulfillig certai coditios, e.g. qt = t t γ, 0 γ < /2, ad G with a certai rate. I Kirch ad Tadjuidje-Kamgaig [6] the asymptotics of the above statistics uder the ull hypothesis ad certai regularity coditios are obtaied. Bt T 2, q coverges i distributio to sup 0<t<, where B is a Browia bridge. qt The other two statistics coverge a.s. to ifiity but such that α T β for some α, β coverges to a Gumbel distributio, which suffices to obtai asymptotic critical values. To obtai tests with asymptotic power oe i these cases, it suffices to show that α β T. For T, it holds α, β, log log /2 ad for T 3, G it holds α 3, β 3, log/g /2, where c d c d c for some costat c > 0. I Kirch ad Tadjuidje-Kamgaig [6] it was show that the above tests have asymptotic power oe for certai fixed alteratives, where Y t, = Y t, usig a uiform law of large umbers. Usig the uiform cetral limit theorem of the previous sectio istead allows us to geeralize this result to a larger class of alteratives, icludig local alteratives such as Y t, = X t + d Z t for some d 0 or X t as i.3 ad Y t, = fy t,, θ + e t with θ θ 0. Local alteratives are ofte cosidered i statistics to get a idea about the sesitivity of the test for small differeces. A.. The chage-poit fulfills k = λ for some 0 < λ <. A. 2. {X t } fulfills 2. as well as t= X t EX = O. A. 3. It holds b EfXp, θ θ= b θ EX 2 for b specified below. Uder local alteratives the estimator θ will typically coverge to θ 0 with a rate depedig o the rate of covergece of the local alterative, i.e. the rate with which d 0 resp. θ θ 0 i the examples above. For a geeral eural etwork it is difficult to quatify these rates due to the highly oliear structure of f. I the classical special case of a mea chage i.e. a trivial eural etwork with H = 0 ad X i = µ + e i, Y i, = µ + d + e i, Assumptio A. 3 reduces to the well kow coditio d 0 but b d. The type of mea chage coditio as i A.3 is typical if the test statistics are based o estimated residuals ad already arise i liear regressio models cf. e.g. Hušková ad Koubkova [5]. For some simulatios cocerig detectability of differet types of chages we refer to Kirch ad Tadjuidje-Kamgaig [6]. The ext theorem shows that the tests correspodig to the statistics i 3.2 have asymptotic power oe uder the above coditios. Theorem 3.. Assume that A. A.3 hold with, for a, log log b =, for b, G for c. log/g, 5
6 4 roofs The it holds a log log /2 T,, b T 2, q, c log/g /2 T 3,, if mi η t η qt > 0 for ay η > 0 ad G, log /G 0, G/ 0. 4 roofs We start with some auxiliary lemmas. Lemma 4.. Let fx, θ, be a eural etwork fulfillig C., x = x,..., x p T. The, for ay θ, θ 2 Θ there exists a costat D > 0 ot depedig o θ, θ 2 or x such that fx, θ fx, θ 2 D θ θ 2 2 where 2 is the Euclidia orm. max x i, i=,...,p roof. A applicatio of the mea value theorem with respect to θ yields for ay θ, θ 2 Θ fx, θ fx, θ 2 = fx, ξ T θ θ 2 for some ξ coθ, the covex hull of Θ. Sice Θ is compact, coθ is compact. By this ad the boudedess of the derivative of ψ we fid D > 0 such that fx, ξ T θ θ 2 θ θ 2 2 fx, ξ 2 D θ θ 2 2 max x i. i=,...,p Lemma 4.2. Let fx, θ, be a eural etwork fulfillig C.. The, there exists R > 0, such that for 0 < R it holds for ay probability measure Q R F L N, F,. L 2 Q 2 Q d where F x = D max i=,...,p x i for some costat D > 0, d = Hp ad fx L 2 Q = f 2 x dqx /2. roof. Deote by N [ ] ɛ, F,. the bracketig umber ad by Mɛ, F,. the packig umber, the it holds cf. Va der Vaart ad Weller [9] N, F,. N [ ] 2, F,., 4. N, F,. M, F,
7 4 roofs Lemma 4. ad Theorem 2.7. i Va der Vaart ad Weller [9] yield for ay > 0 N [ ] 2 F L 2 Q, F,. L 2 Q N, F,. 2. Hece by 4. w.l.o.g. F L 2 Q < N, F, L 2 Q N, F, 2 F L 2 Q for which 4.2 implies N, F,. 2 N, F L 2 Q F F,. 2 M, L 2 Q F F,. 2 L 2 Q 3R F d L 2 Q, for ay 0 < R, where F = {fθ, ; θ B0, R} for a suitable R > 0, B0, R is the ball aroud 0 of radius R i R d. The last lie follows from Exercise 6, page 94, of Va der Vaart ad Weller [9]. roof of Theorem 2.. The assertio follows from Lemma 2. of Arcoes ad Yu []. The statemet there is give for the miimal evelope fuctio but the proof shows that it remais true for ay evelope fuctio F. Furthermore it is sufficiet if their coditio 2.0 holds for all 0 < 0 for some 0 > 0. I our case we cosider F as i Lemma 4.. It remais to check the coditios of Lemma 2. of Arcoes ad Yu []. Let p = miτ, ν, the their Coditio 2.3 holds by E X t p < accordig to C.2. Coditio 2.4 holds by C.2, 2.0 by C.3 ad 2. for small by Lemma 4.2. roof of Theorem 3.. Ŝ k = k j=p+ By 3., A. ad A.2 it holds X j fx j, θ k = X i + k EfX p, θ θ= θ b + O sup i=p+ θ Θ = λ EX EfX p, θ θ= θ b + O By A.3 this implies log log /2 T, λ log log k i=p+ fx i, θ EfX p, θ EX EfX p, θ θ= b θ + O log log /2 Sice by A. ad mi η t η qt > c for ay η > 0 it holds qk c for some c > 0, which together with A.3 implies T 2, q λ EX EfX p, θ θ= b θ + O. Similarly Ŝk Ŝk G = G EX EfX p, θ θ= θ b + O G hece by A.3 log/g /2 T 3, G G log/g EX EfX p, θ θ= b θ +O log/g /2.. 7
8 Refereces Refereces [] Arcoes, M. A. ad Yu, B. Cetral limit theorems for empirical ad U-processes of statioary mixig sequeces. Ecoometrica, 7:48 7, 994. [2] Csörgő, M., ad Horváth, L. Limit Theorems i Chage-oit Aalysis. Wiley, Chichester, 997. [3] Dudley, R.M. Cetral limit theorem for empirical processes. A. robab., 6: , 978. [4] Frake, J. ad Mabouba, D. Estimatig market risk with eural etworks. Statist. Decisios, 30:63 82, [5] Hušková, M. ad Koubková, A. Moitorig jump chages i liear models. J. Statist. Res., 39:59 78, [6] Kirch, C. ad Tadjuidje K., J.. Testig for parameter stability i oliear autoregressive models. reprit, 20. [7] Mey, S.. ad Tweedie, R.L.. Markov Chais ad Stochastic Stability. Spriger, Lodo, 993. [8] Stockis, J.-., Frake, J., ad Tadjuidje K., J. O geometric ergodicity of CHARME models. J. Time Ser. Aal., 3:4 52, 200. [9] Va der Vaart, A. W. ad Weller, J. A. Weak covergece ad empirical processes. Spriger, New York, 996. [0] White, H. Coectioist oparametric regressio: Multilayer feedforward etworks ca lear arbitrary mappigs. Neural Networks, 3: ,
A uniform central limit theorem for neural network based autoregressive processes with applications to change-point analysis
A uniform central limit theorem for neural network based autoregressive processes with applications to change-point analysis Claudia Kirch Joseph Tadjuidje Kamgaing March 6, 20 Abstract We consider an
More informationEmpirical Processes: Glivenko Cantelli Theorems
Empirical Processes: Gliveko Catelli Theorems Mouliath Baerjee Jue 6, 200 Gliveko Catelli classes of fuctios The reader is referred to Chapter.6 of Weller s Torgo otes, Chapter??? of VDVW ad Chapter 8.3
More informationA RANK STATISTIC FOR NON-PARAMETRIC K-SAMPLE AND CHANGE POINT PROBLEMS
J. Japa Statist. Soc. Vol. 41 No. 1 2011 67 73 A RANK STATISTIC FOR NON-PARAMETRIC K-SAMPLE AND CHANGE POINT PROBLEMS Yoichi Nishiyama* We cosider k-sample ad chage poit problems for idepedet data i a
More informationEstimation of the essential supremum of a regression function
Estimatio of the essetial supremum of a regressio fuctio Michael ohler, Adam rzyżak 2, ad Harro Walk 3 Fachbereich Mathematik, Techische Uiversität Darmstadt, Schlossgartestr. 7, 64289 Darmstadt, Germay,
More informationConvergence of random variables. (telegram style notes) P.J.C. Spreij
Covergece of radom variables (telegram style otes).j.c. Spreij this versio: September 6, 2005 Itroductio As we kow, radom variables are by defiitio measurable fuctios o some uderlyig measurable space
More informationResampling Methods. X (1/2), i.e., Pr (X i m) = 1/2. We order the data: X (1) X (2) X (n). Define the sample median: ( n.
Jauary 1, 2019 Resamplig Methods Motivatio We have so may estimators with the property θ θ d N 0, σ 2 We ca also write θ a N θ, σ 2 /, where a meas approximately distributed as Oce we have a cosistet estimator
More informationSequences and Series of Functions
Chapter 6 Sequeces ad Series of Fuctios 6.1. Covergece of a Sequece of Fuctios Poitwise Covergece. Defiitio 6.1. Let, for each N, fuctio f : A R be defied. If, for each x A, the sequece (f (x)) coverges
More informationDefinition 4.2. (a) A sequence {x n } in a Banach space X is a basis for X if. unique scalars a n (x) such that x = n. a n (x) x n. (4.
4. BASES I BAACH SPACES 39 4. BASES I BAACH SPACES Sice a Baach space X is a vector space, it must possess a Hamel, or vector space, basis, i.e., a subset {x γ } γ Γ whose fiite liear spa is all of X ad
More informationIt is often useful to approximate complicated functions using simpler ones. We consider the task of approximating a function by a polynomial.
Taylor Polyomials ad Taylor Series It is ofte useful to approximate complicated fuctios usig simpler oes We cosider the task of approximatig a fuctio by a polyomial If f is at least -times differetiable
More information32 estimating the cumulative distribution function
32 estimatig the cumulative distributio fuctio 4.6 types of cofidece itervals/bads Let F be a class of distributio fuctios F ad let θ be some quatity of iterest, such as the mea of F or the whole fuctio
More informationLecture 27: Optimal Estimators and Functional Delta Method
Stat210B: Theoretical Statistics Lecture Date: April 19, 2007 Lecture 27: Optimal Estimators ad Fuctioal Delta Method Lecturer: Michael I. Jorda Scribe: Guilherme V. Rocha 1 Achievig Optimal Estimators
More informationECONOMETRIC THEORY. MODULE XIII Lecture - 34 Asymptotic Theory and Stochastic Regressors
ECONOMETRIC THEORY MODULE XIII Lecture - 34 Asymptotic Theory ad Stochastic Regressors Dr. Shalabh Departmet of Mathematics ad Statistics Idia Istitute of Techology Kapur Asymptotic theory The asymptotic
More informationRates of Convergence by Moduli of Continuity
Rates of Covergece by Moduli of Cotiuity Joh Duchi: Notes for Statistics 300b March, 017 1 Itroductio I this ote, we give a presetatio showig the importace, ad relatioship betwee, the modulis of cotiuity
More informationECE 901 Lecture 12: Complexity Regularization and the Squared Loss
ECE 90 Lecture : Complexity Regularizatio ad the Squared Loss R. Nowak 5/7/009 I the previous lectures we made use of the Cheroff/Hoeffdig bouds for our aalysis of classifier errors. Hoeffdig s iequality
More informationAppendix to: Hypothesis Testing for Multiple Mean and Correlation Curves with Functional Data
Appedix to: Hypothesis Testig for Multiple Mea ad Correlatio Curves with Fuctioal Data Ao Yua 1, Hog-Bi Fag 1, Haiou Li 1, Coli O. Wu, Mig T. Ta 1, 1 Departmet of Biostatistics, Bioiformatics ad Biomathematics,
More informationLecture 33: Bootstrap
Lecture 33: ootstrap Motivatio To evaluate ad compare differet estimators, we eed cosistet estimators of variaces or asymptotic variaces of estimators. This is also importat for hypothesis testig ad cofidece
More informationIntroduction to Extreme Value Theory Laurens de Haan, ISM Japan, Erasmus University Rotterdam, NL University of Lisbon, PT
Itroductio to Extreme Value Theory Laures de Haa, ISM Japa, 202 Itroductio to Extreme Value Theory Laures de Haa Erasmus Uiversity Rotterdam, NL Uiversity of Lisbo, PT Itroductio to Extreme Value Theory
More informationChapter 7 Isoperimetric problem
Chapter 7 Isoperimetric problem Recall that the isoperimetric problem (see the itroductio its coectio with ido s proble) is oe of the most classical problem of a shape optimizatio. It ca be formulated
More informationb i u x i U a i j u x i u x j
M ath 5 2 7 Fall 2 0 0 9 L ecture 1 9 N ov. 1 6, 2 0 0 9 ) S ecod- Order Elliptic Equatios: Weak S olutios 1. Defiitios. I this ad the followig two lectures we will study the boudary value problem Here
More informationAsymptotic distribution of products of sums of independent random variables
Proc. Idia Acad. Sci. Math. Sci. Vol. 3, No., May 03, pp. 83 9. c Idia Academy of Scieces Asymptotic distributio of products of sums of idepedet radom variables YANLING WANG, SUXIA YAO ad HONGXIA DU ollege
More informationLecture 19: Convergence
Lecture 19: Covergece Asymptotic approach I statistical aalysis or iferece, a key to the success of fidig a good procedure is beig able to fid some momets ad/or distributios of various statistics. I may
More informationLet us give one more example of MLE. Example 3. The uniform distribution U[0, θ] on the interval [0, θ] has p.d.f.
Lecture 5 Let us give oe more example of MLE. Example 3. The uiform distributio U[0, ] o the iterval [0, ] has p.d.f. { 1 f(x =, 0 x, 0, otherwise The likelihood fuctio ϕ( = f(x i = 1 I(X 1,..., X [0,
More informationFall 2013 MTH431/531 Real analysis Section Notes
Fall 013 MTH431/531 Real aalysis Sectio 8.1-8. Notes Yi Su 013.11.1 1. Defiitio of uiform covergece. We look at a sequece of fuctios f (x) ad study the coverget property. Notice we have two parameters
More informationACO Comprehensive Exam 9 October 2007 Student code A. 1. Graph Theory
1. Graph Theory Prove that there exist o simple plaar triagulatio T ad two distict adjacet vertices x, y V (T ) such that x ad y are the oly vertices of T of odd degree. Do ot use the Four-Color Theorem.
More informationA survey on penalized empirical risk minimization Sara A. van de Geer
A survey o pealized empirical risk miimizatio Sara A. va de Geer We address the questio how to choose the pealty i empirical risk miimizatio. Roughly speakig, this pealty should be a good boud for the
More informationBull. Korean Math. Soc. 36 (1999), No. 3, pp. 451{457 THE STRONG CONSISTENCY OF NONLINEAR REGRESSION QUANTILES ESTIMATORS Seung Hoe Choi and Hae Kyung
Bull. Korea Math. Soc. 36 (999), No. 3, pp. 45{457 THE STRONG CONSISTENCY OF NONLINEAR REGRESSION QUANTILES ESTIMATORS Abstract. This paper provides suciet coditios which esure the strog cosistecy of regressio
More informationSelf-normalized deviation inequalities with application to t-statistic
Self-ormalized deviatio iequalities with applicatio to t-statistic Xiequa Fa Ceter for Applied Mathematics, Tiaji Uiversity, 30007 Tiaji, Chia Abstract Let ξ i i 1 be a sequece of idepedet ad symmetric
More informationKernel density estimator
Jauary, 07 NONPARAMETRIC ERNEL DENSITY ESTIMATION I this lecture, we discuss kerel estimatio of probability desity fuctios PDF Noparametric desity estimatio is oe of the cetral problems i statistics I
More informationA statistical method to determine sample size to estimate characteristic value of soil parameters
A statistical method to determie sample size to estimate characteristic value of soil parameters Y. Hojo, B. Setiawa 2 ad M. Suzuki 3 Abstract Sample size is a importat factor to be cosidered i determiig
More informationBerry-Esseen bounds for self-normalized martingales
Berry-Essee bouds for self-ormalized martigales Xiequa Fa a, Qi-Ma Shao b a Ceter for Applied Mathematics, Tiaji Uiversity, Tiaji 30007, Chia b Departmet of Statistics, The Chiese Uiversity of Hog Kog,
More informationOn Orlicz N-frames. 1 Introduction. Renu Chugh 1,, Shashank Goel 2
Joural of Advaced Research i Pure Mathematics Olie ISSN: 1943-2380 Vol. 3, Issue. 1, 2010, pp. 104-110 doi: 10.5373/jarpm.473.061810 O Orlicz N-frames Reu Chugh 1,, Shashak Goel 2 1 Departmet of Mathematics,
More informationKolmogorov-Smirnov type Tests for Local Gaussianity in High-Frequency Data
Proceedigs 59th ISI World Statistics Cogress, 5-30 August 013, Hog Kog (Sessio STS046) p.09 Kolmogorov-Smirov type Tests for Local Gaussiaity i High-Frequecy Data George Tauche, Duke Uiversity Viktor Todorov,
More informationSolutions to home assignments (sketches)
Matematiska Istitutioe Peter Kumli 26th May 2004 TMA401 Fuctioal Aalysis MAN670 Applied Fuctioal Aalysis 4th quarter 2003/2004 All documet cocerig the course ca be foud o the course home page: http://www.math.chalmers.se/math/grudutb/cth/tma401/
More informationQuantile regression with multilayer perceptrons.
Quatile regressio with multilayer perceptros. S.-F. Dimby ad J. Rykiewicz Uiversite Paris 1 - SAMM 90 Rue de Tolbiac, 75013 Paris - Frace Abstract. We cosider oliear quatile regressio ivolvig multilayer
More information62. Power series Definition 16. (Power series) Given a sequence {c n }, the series. c n x n = c 0 + c 1 x + c 2 x 2 + c 3 x 3 +
62. Power series Defiitio 16. (Power series) Give a sequece {c }, the series c x = c 0 + c 1 x + c 2 x 2 + c 3 x 3 + is called a power series i the variable x. The umbers c are called the coefficiets of
More informationApproximation by Superpositions of a Sigmoidal Function
Zeitschrift für Aalysis ud ihre Aweduge Joural for Aalysis ad its Applicatios Volume 22 (2003, No. 2, 463 470 Approximatio by Superpositios of a Sigmoidal Fuctio G. Lewicki ad G. Mario Abstract. We geeralize
More informationEntropy Rates and Asymptotic Equipartition
Chapter 29 Etropy Rates ad Asymptotic Equipartitio Sectio 29. itroduces the etropy rate the asymptotic etropy per time-step of a stochastic process ad shows that it is well-defied; ad similarly for iformatio,
More informationAN EXTENSION OF SIMONS INEQUALITY AND APPLICATIONS. Robert DEVILLE and Catherine FINET
2001 vol. XIV, um. 1, 95-104 ISSN 1139-1138 AN EXTENSION OF SIMONS INEQUALITY AND APPLICATIONS Robert DEVILLE ad Catherie FINET Abstract This article is devoted to a extesio of Simos iequality. As a cosequece,
More informationOptimally Sparse SVMs
A. Proof of Lemma 3. We here prove a lower boud o the umber of support vectors to achieve geeralizatio bouds of the form which we cosider. Importatly, this result holds ot oly for liear classifiers, but
More informationProduct measures, Tonelli s and Fubini s theorems For use in MAT3400/4400, autumn 2014 Nadia S. Larsen. Version of 13 October 2014.
Product measures, Toelli s ad Fubii s theorems For use i MAT3400/4400, autum 2014 Nadia S. Larse Versio of 13 October 2014. 1. Costructio of the product measure The purpose of these otes is to preset the
More informationMASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.436J/15.085J Fall 2008 Lecture 19 11/17/2008 LAWS OF LARGE NUMBERS II THE STRONG LAW OF LARGE NUMBERS
MASSACHUSTTS INSTITUT OF TCHNOLOGY 6.436J/5.085J Fall 2008 Lecture 9 /7/2008 LAWS OF LARG NUMBRS II Cotets. The strog law of large umbers 2. The Cheroff boud TH STRONG LAW OF LARG NUMBRS While the weak
More information10-701/ Machine Learning Mid-term Exam Solution
0-70/5-78 Machie Learig Mid-term Exam Solutio Your Name: Your Adrew ID: True or False (Give oe setece explaatio) (20%). (F) For a cotiuous radom variable x ad its probability distributio fuctio p(x), it
More informationPreponderantly increasing/decreasing data in regression analysis
Croatia Operatioal Research Review 269 CRORR 7(2016), 269 276 Prepoderatly icreasig/decreasig data i regressio aalysis Darija Marković 1, 1 Departmet of Mathematics, J. J. Strossmayer Uiversity of Osijek,
More information5.1 A mutual information bound based on metric entropy
Chapter 5 Global Fao Method I this chapter, we exted the techiques of Chapter 2.4 o Fao s method the local Fao method) to a more global costructio. I particular, we show that, rather tha costructig a local
More informationAlgebra of Least Squares
October 19, 2018 Algebra of Least Squares Geometry of Least Squares Recall that out data is like a table [Y X] where Y collects observatios o the depedet variable Y ad X collects observatios o the k-dimesioal
More informationChapter 6 Principles of Data Reduction
Chapter 6 for BST 695: Special Topics i Statistical Theory. Kui Zhag, 0 Chapter 6 Priciples of Data Reductio Sectio 6. Itroductio Goal: To summarize or reduce the data X, X,, X to get iformatio about a
More informationOn forward improvement iteration for stopping problems
O forward improvemet iteratio for stoppig problems Mathematical Istitute, Uiversity of Kiel, Ludewig-Mey-Str. 4, D-24098 Kiel, Germay irle@math.ui-iel.de Albrecht Irle Abstract. We cosider the optimal
More informationNotes 19 : Martingale CLT
Notes 9 : Martigale CLT Math 733-734: Theory of Probability Lecturer: Sebastie Roch Refereces: [Bil95, Chapter 35], [Roc, Chapter 3]. Sice we have ot ecoutered weak covergece i some time, we first recall
More informationPermutation principles for the change analysis of stochastic processes under strong invariance
Permutatio priciples for the chage aalysis of stochastic processes uder strog ivariace Claudia Kirch, Josef Steiebach September 2, 2004 Abstract Approximatios of the critical values for chage-poit tests
More informationLecture 8: Convergence of transformations and law of large numbers
Lecture 8: Covergece of trasformatios ad law of large umbers Trasformatio ad covergece Trasformatio is a importat tool i statistics. If X coverges to X i some sese, we ofte eed to check whether g(x ) coverges
More information6.3 Testing Series With Positive Terms
6.3. TESTING SERIES WITH POSITIVE TERMS 307 6.3 Testig Series With Positive Terms 6.3. Review of what is kow up to ow I theory, testig a series a i for covergece amouts to fidig the i= sequece of partial
More informationProblem Set 4 Due Oct, 12
EE226: Radom Processes i Systems Lecturer: Jea C. Walrad Problem Set 4 Due Oct, 12 Fall 06 GSI: Assae Gueye This problem set essetially reviews detectio theory ad hypothesis testig ad some basic otios
More informationSummary and Discussion on Simultaneous Analysis of Lasso and Dantzig Selector
Summary ad Discussio o Simultaeous Aalysis of Lasso ad Datzig Selector STAT732, Sprig 28 Duzhe Wag May 4, 28 Abstract This is a discussio o the work i Bickel, Ritov ad Tsybakov (29). We begi with a short
More information1 of 7 7/16/2009 6:06 AM Virtual Laboratories > 6. Radom Samples > 1 2 3 4 5 6 7 6. Order Statistics Defiitios Suppose agai that we have a basic radom experimet, ad that X is a real-valued radom variable
More informationlim za n n = z lim a n n.
Lecture 6 Sequeces ad Series Defiitio 1 By a sequece i a set A, we mea a mappig f : N A. It is customary to deote a sequece f by {s } where, s := f(). A sequece {z } of (complex) umbers is said to be coverget
More informationAdvanced Analysis. Min Yan Department of Mathematics Hong Kong University of Science and Technology
Advaced Aalysis Mi Ya Departmet of Mathematics Hog Kog Uiversity of Sciece ad Techology September 3, 009 Cotets Limit ad Cotiuity 7 Limit of Sequece 8 Defiitio 8 Property 3 3 Ifiity ad Ifiitesimal 8 4
More informationChapter 3. Strong convergence. 3.1 Definition of almost sure convergence
Chapter 3 Strog covergece As poited out i the Chapter 2, there are multiple ways to defie the otio of covergece of a sequece of radom variables. That chapter defied covergece i probability, covergece i
More informationREGRESSION WITH QUADRATIC LOSS
REGRESSION WITH QUADRATIC LOSS MAXIM RAGINSKY Regressio with quadratic loss is aother basic problem studied i statistical learig theory. We have a radom couple Z = X, Y ), where, as before, X is a R d
More informationMath Solutions to homework 6
Math 175 - Solutios to homework 6 Cédric De Groote November 16, 2017 Problem 1 (8.11 i the book): Let K be a compact Hermitia operator o a Hilbert space H ad let the kerel of K be {0}. Show that there
More information4. Partial Sums and the Central Limit Theorem
1 of 10 7/16/2009 6:05 AM Virtual Laboratories > 6. Radom Samples > 1 2 3 4 5 6 7 4. Partial Sums ad the Cetral Limit Theorem The cetral limit theorem ad the law of large umbers are the two fudametal theorems
More informationsin(n) + 2 cos(2n) n 3/2 3 sin(n) 2cos(2n) n 3/2 a n =
60. Ratio ad root tests 60.1. Absolutely coverget series. Defiitio 13. (Absolute covergece) A series a is called absolutely coverget if the series of absolute values a is coverget. The absolute covergece
More informationLECTURE 8: ASYMPTOTICS I
LECTURE 8: ASYMPTOTICS I We are iterested i the properties of estimators as. Cosider a sequece of radom variables {, X 1}. N. M. Kiefer, Corell Uiversity, Ecoomics 60 1 Defiitio: (Weak covergece) A sequece
More informationAn Introduction to Asymptotic Theory
A Itroductio to Asymptotic Theory Pig Yu School of Ecoomics ad Fiace The Uiversity of Hog Kog Pig Yu (HKU) Asymptotic Theory 1 / 20 Five Weapos i Asymptotic Theory Five Weapos i Asymptotic Theory Pig Yu
More informationEmpirical Process Theory and Oracle Inequalities
Stat 928: Statistical Learig Theory Lecture: 10 Empirical Process Theory ad Oracle Iequalities Istructor: Sham Kakade 1 Risk vs Risk See Lecture 0 for a discussio o termiology. 2 The Uio Boud / Boferoi
More informationBoundaries and the James theorem
Boudaries ad the James theorem L. Vesely 1. Itroductio The followig theorem is importat ad well kow. All spaces cosidered here are real ormed or Baach spaces. Give a ormed space X, we deote by B X ad S
More informationSlide Set 13 Linear Model with Endogenous Regressors and the GMM estimator
Slide Set 13 Liear Model with Edogeous Regressors ad the GMM estimator Pietro Coretto pcoretto@uisa.it Ecoometrics Master i Ecoomics ad Fiace (MEF) Uiversità degli Studi di Napoli Federico II Versio: Friday
More informationChapter 6 Infinite Series
Chapter 6 Ifiite Series I the previous chapter we cosidered itegrals which were improper i the sese that the iterval of itegratio was ubouded. I this chapter we are goig to discuss a topic which is somewhat
More informationOn the detection of changes in autoregressive time series, II. Resampling procedures 1
O the detectio of chages i autoregressive time series, II. Resamplig procedures Marie Hušková, Claudia Kirch 2, Zuzaa Prášková 3, Josef Steiebach 4 Charles Uiversity of Prague, Departmet of Statistics,
More informationAda Boost, Risk Bounds, Concentration Inequalities. 1 AdaBoost and Estimates of Conditional Probabilities
CS8B/Stat4B Sprig 008) Statistical Learig Theory Lecture: Ada Boost, Risk Bouds, Cocetratio Iequalities Lecturer: Peter Bartlett Scribe: Subhrasu Maji AdaBoost ad Estimates of Coditioal Probabilities We
More informationMASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 21 11/27/2013
MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 21 11/27/2013 Fuctioal Law of Large Numbers. Costructio of the Wieer Measure Cotet. 1. Additioal techical results o weak covergece
More information7.1 Convergence of sequences of random variables
Chapter 7 Limit Theorems Throughout this sectio we will assume a probability space (, F, P), i which is defied a ifiite sequece of radom variables (X ) ad a radom variable X. The fact that for every ifiite
More information1 = δ2 (0, ), Y Y n nδ. , T n = Y Y n n. ( U n,k + X ) ( f U n,k + Y ) n 2n f U n,k + θ Y ) 2 E X1 2 X1
8. The cetral limit theorems 8.1. The cetral limit theorem for i.i.d. sequeces. ecall that C ( is N -separatig. Theorem 8.1. Let X 1, X,... be i.i.d. radom variables with EX 1 = ad EX 1 = σ (,. Suppose
More informationExpectation and Variance of a random variable
Chapter 11 Expectatio ad Variace of a radom variable The aim of this lecture is to defie ad itroduce mathematical Expectatio ad variace of a fuctio of discrete & cotiuous radom variables ad the distributio
More informationDiscrete Mathematics for CS Spring 2007 Luca Trevisan Lecture 22
CS 70 Discrete Mathematics for CS Sprig 2007 Luca Trevisa Lecture 22 Aother Importat Distributio The Geometric Distributio Questio: A biased coi with Heads probability p is tossed repeatedly util the first
More informationAn alternative proof of a theorem of Aldous. concerning convergence in distribution for martingales.
A alterative proof of a theorem of Aldous cocerig covergece i distributio for martigales. Maurizio Pratelli Dipartimeto di Matematica, Uiversità di Pisa. Via Buoarroti 2. I-56127 Pisa, Italy e-mail: pratelli@dm.uipi.it
More informationMAT1026 Calculus II Basic Convergence Tests for Series
MAT026 Calculus II Basic Covergece Tests for Series Egi MERMUT 202.03.08 Dokuz Eylül Uiversity Faculty of Sciece Departmet of Mathematics İzmir/TURKEY Cotets Mootoe Covergece Theorem 2 2 Series of Real
More informationChapter 13: Tests of Hypothesis Section 13.1 Introduction
Chapter 13: Tests of Hypothesis Sectio 13.1 Itroductio RECAP: Chapter 1 discussed the Likelihood Ratio Method as a geeral approach to fid good test procedures. Testig for the Normal Mea Example, discussed
More informationRandom Variables, Sampling and Estimation
Chapter 1 Radom Variables, Samplig ad Estimatio 1.1 Itroductio This chapter will cover the most importat basic statistical theory you eed i order to uderstad the ecoometric material that will be comig
More informationPRELIM PROBLEM SOLUTIONS
PRELIM PROBLEM SOLUTIONS THE GRAD STUDENTS + KEN Cotets. Complex Aalysis Practice Problems 2. 2. Real Aalysis Practice Problems 2. 4 3. Algebra Practice Problems 2. 8. Complex Aalysis Practice Problems
More informationTESTING FOR THE BUFFERED AUTOREGRESSIVE PROCESSES (SUPPLEMENTARY MATERIAL)
TESTING FOR THE BUFFERED AUTOREGRESSIVE PROCESSES SUPPLEMENTARY MATERIAL) By Ke Zhu, Philip L.H. Yu ad Wai Keug Li Chiese Academy of Scieces ad Uiversity of Hog Kog APPENDIX: PROOFS I this appedix, we
More informationStochastic Simulation
Stochastic Simulatio 1 Itroductio Readig Assigmet: Read Chapter 1 of text. We shall itroduce may of the key issues to be discussed i this course via a couple of model problems. Model Problem 1 (Jackso
More informationLecture 3: August 31
36-705: Itermediate Statistics Fall 018 Lecturer: Siva Balakrisha Lecture 3: August 31 This lecture will be mostly a summary of other useful expoetial tail bouds We will ot prove ay of these i lecture,
More informationRiesz-Fischer Sequences and Lower Frame Bounds
Zeitschrift für Aalysis ud ihre Aweduge Joural for Aalysis ad its Applicatios Volume 1 (00), No., 305 314 Riesz-Fischer Sequeces ad Lower Frame Bouds P. Casazza, O. Christese, S. Li ad A. Lider Abstract.
More information1 Convergence in Probability and the Weak Law of Large Numbers
36-752 Advaced Probability Overview Sprig 2018 8. Covergece Cocepts: i Probability, i L p ad Almost Surely Istructor: Alessadro Rialdo Associated readig: Sec 2.4, 2.5, ad 4.11 of Ash ad Doléas-Dade; Sec
More information1 Inferential Methods for Correlation and Regression Analysis
1 Iferetial Methods for Correlatio ad Regressio Aalysis I the chapter o Correlatio ad Regressio Aalysis tools for describig bivariate cotiuous data were itroduced. The sample Pearso Correlatio Coefficiet
More informationDiscrete Mathematics for CS Spring 2008 David Wagner Note 22
CS 70 Discrete Mathematics for CS Sprig 2008 David Wager Note 22 I.I.D. Radom Variables Estimatig the bias of a coi Questio: We wat to estimate the proportio p of Democrats i the US populatio, by takig
More informationApplication to Random Graphs
A Applicatio to Radom Graphs Brachig processes have a umber of iterestig ad importat applicatios. We shall cosider oe of the most famous of them, the Erdős-Réyi radom graph theory. 1 Defiitio A.1. Let
More informationIntroduction to Probability. Ariel Yadin
Itroductio to robability Ariel Yadi Lecture 2 *** Ja. 7 ***. Covergece of Radom Variables As i the case of sequeces of umbers, we would like to talk about covergece of radom variables. There are may ways
More informationLONG SNAKES IN POWERS OF THE COMPLETE GRAPH WITH AN ODD NUMBER OF VERTICES
J Lodo Math Soc (2 50, (1994, 465 476 LONG SNAKES IN POWERS OF THE COMPLETE GRAPH WITH AN ODD NUMBER OF VERTICES Jerzy Wojciechowski Abstract I [5] Abbott ad Katchalski ask if there exists a costat c >
More informationRegression with an Evaporating Logarithmic Trend
Regressio with a Evaporatig Logarithmic Tred Peter C. B. Phillips Cowles Foudatio, Yale Uiversity, Uiversity of Aucklad & Uiversity of York ad Yixiao Su Departmet of Ecoomics Yale Uiversity October 5,
More information1 The Haar functions and the Brownian motion
1 The Haar fuctios ad the Browia motio 1.1 The Haar fuctios ad their completeess The Haar fuctios The basic Haar fuctio is 1 if x < 1/2, ψx) = 1 if 1/2 x < 1, otherwise. 1.1) It has mea zero 1 ψx)dx =,
More informationStochastic Matrices in a Finite Field
Stochastic Matrices i a Fiite Field Abstract: I this project we will explore the properties of stochastic matrices i both the real ad the fiite fields. We first explore what properties 2 2 stochastic matrices
More informationAxioms of Measure Theory
MATH 532 Axioms of Measure Theory Dr. Neal, WKU I. The Space Throughout the course, we shall let X deote a geeric o-empty set. I geeral, we shall ot assume that ay algebraic structure exists o X so that
More informationki, X(n) lj (n) = (ϱ (n) ij ) 1 i,j d.
APPLICATIONES MATHEMATICAE 22,2 (1994), pp. 193 200 M. WIŚNIEWSKI (Kielce) EXTREME ORDER STATISTICS IN AN EQUALLY CORRELATED GAUSSIAN ARRAY Abstract. This paper cotais the results cocerig the wea covergece
More informationA Note on the Kolmogorov-Feller Weak Law of Large Numbers
Joural of Mathematical Research with Applicatios Mar., 015, Vol. 35, No., pp. 3 8 DOI:10.3770/j.iss:095-651.015.0.013 Http://jmre.dlut.edu.c A Note o the Kolmogorov-Feller Weak Law of Large Numbers Yachu
More informationRandom Walks on Discrete and Continuous Circles. by Jeffrey S. Rosenthal School of Mathematics, University of Minnesota, Minneapolis, MN, U.S.A.
Radom Walks o Discrete ad Cotiuous Circles by Jeffrey S. Rosethal School of Mathematics, Uiversity of Miesota, Mieapolis, MN, U.S.A. 55455 (Appeared i Joural of Applied Probability 30 (1993), 780 789.)
More informationMASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/15.070J Fall 2013 Lecture 6 9/23/2013. Brownian motion. Introduction
MASSACHUSETTS INSTITUTE OF TECHNOLOGY 6.265/5.070J Fall 203 Lecture 6 9/23/203 Browia motio. Itroductio Cotet.. A heuristic costructio of a Browia motio from a radom walk. 2. Defiitio ad basic properties
More information7.1 Convergence of sequences of random variables
Chapter 7 Limit theorems Throughout this sectio we will assume a probability space (Ω, F, P), i which is defied a ifiite sequece of radom variables (X ) ad a radom variable X. The fact that for every ifiite
More informationGamma Distribution and Gamma Approximation
Gamma Distributio ad Gamma Approimatio Xiaomig Zeg a Fuhua (Frak Cheg b a Xiame Uiversity, Xiame 365, Chia mzeg@jigia.mu.edu.c b Uiversity of Ketucky, Leigto, Ketucky 456-46, USA cheg@cs.uky.edu Abstract
More information6. Kalman filter implementation for linear algebraic equations. Karhunen-Loeve decomposition
6. Kalma filter implemetatio for liear algebraic equatios. Karhue-Loeve decompositio 6.1. Solvable liear algebraic systems. Probabilistic iterpretatio. Let A be a quadratic matrix (ot obligatory osigular.
More information