1 / 24

Stochastic models - time series.

Stochastic models - time series. Random process . an infinite collection of consistent distributions probabilities exist Random function . a family of random variables, e.g. {Y(t), t in Z}. Specified if given

Antony
Télécharger la présentation

Stochastic models - time series.

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Stochastic models - time series. Random process. an infinite collection of consistent distributions probabilities exist Random function. a family of random variables, e.g. {Y(t), t in Z}

  2. Specified if given F(y1,...,yn;t1 ,...,tn ) = Prob{Y(t1)y1,...,Y(tn )yn } that are symmetric F(y;t) = F(y;t),  a permutation compatible F(y1 ,...,ym ,,...,;t1,...,tm,tm+1,...,tn} = F(y1,...,ym;t1,...,tm)

  3. Finite dimensional distributions First-order F(y;t) = Prob{Y(t)  t} Second-order F(y1,y2;t1,t2) = Prob{Y(t1)  y1 and Y(t2)  y2} and so on

  4. Other methods i) Y(t;), : random variable ii) urn model iii) probability on function space iv) analytic formula Y(t) =  cos(t + ) : fixed : uniform on (-,]

  5. There may be densities The Y(t) may be discrete, angles, proportions, ... Kolmogorov extension theorem. To specify a stochastic process give the distribution of any finite subset {Y(1),...,Y(n)} in a consistent way,  in A

  6. Moment functions. Mean function cY(t) = E{Y(t)} =  y dF(y;t) =  y f(y;t) dy if continuous =  yjf(yj; t) if discrete E{1Y1(t) + 2Y2(t)} =1c1(t) +2c2(t) vector-valued case mean level - signal plus noise: S(t) + (t) S(.): fixed

  7. Second-moments. autocovariance function cYY(s,t) = cov{Y(s),Y(t)} = E{Y(s)Y(t)} - E{Y(s)}E{Y(t)} non-negative definite jkcYY(tj , tk )  0 scalars  crosscovariance function c12(s,t) = cov{Y1(s),Y2(t)}

  8. Stationarity. Joint distributions, {Y(t+u1),...,Y(t+uk-1),Y(t)}, do not depend on t for k=1,2,... Often reasonable in practice - for some time stretches Replaces "identically distributed"

  9. mean E{Y(t)} = cY for t in Z autocovariance function cov{Y(t+u),Y(t)} = cYY(u) t,u in Z u: lag = E{Y(t+u)Y(t)} if mean 0 autocorrelation function(u) = corr{Y(t+u),Y(t)}, |(u)|  1 crosscovariance function cov{X(t+u),Y(t)} = cXY(u)

  10. joint density Prob{x < Y(t+u) < x+dx and y < Y(t) < y+ dy} = f(x,y|u) dxdy

  11. Some useful modelsChatfield notation Purely random / white noise often mean 0 Building block

  12. Random walk not stationary

  13. (*)

  14. Moving average, MA(q) From (*) stationary

  15. MA(1) 0=1 1 = -.7

  16. Backward shift operator Linear process. Need convergence condition

  17. autoregressive process, AR(p) first-order, AR(1) Markov * Linear process For convergence/stationarity

  18. a.c.f. From (*) p.a.c.f. corr{Y(t),Y(t-m)|Y(t-1),...,Y(t-m+1)} linearly = 0 for m  p when Y is AR(p)

  19. In general case, Useful for prediction

  20. ARMA(p,q)

  21. ARIMA(p,d,q).

  22. Some series and acf’s

  23. Yule-Walker equations for AR(p). Correlate, with Xt-k, each side of

  24. Cumulants. multilinear functional 0 if some subset of variantes independent of rest 0 of order > 2 for normal normal is determined by its moments

More Related