Search

Abstract: A new likelihood based AR approximation is given for ARMA models. The usual algorithms for the computation of the likelihood of an ARMA model require flops per function evaluation. Using our new approximation, an algorithm is developed which requires only flops in repeated likelihood evaluations. In most cases, the new algorithm gives results identical to or very close to the exact maximum likelihood estimate (MLE). This algorithm is easily implemented in high level quantitative programming environments (QPEs) such as Mathematica, MatLab and R. In order to obtain reasonable speed, previous ARMA maximum likelihood algorithms are usually implemented in C or some other machine efficient language. With our algorithm it is easy to do maximum likelihood estimation for long time series directly in the QPE of your choice. The new algorithm is extended to obtain the MLE for the mean parameter. Simulation experiments which illustrate the effectiveness of the new algorithm are discussed. Mathematica and R packages which implement the algorithm discussed in this paper are available [McLeod, A.I., Zhang, Y., 2007. Online supplements to “Faster ARMA Maximum Likelihood Estimation”, http://www.stats.uwo.ca/faculty/aim/2007/faster/ ]. Based on these package implementations, it is expected that the interested researcher would be able to implement this algorithm in other QPEs. [Copyright &y& Elsevier]

This paper proposes a bootstrap test for testing the null hypothesis that a time series is stationary against the alternative hypothesis that it is integrated of order one. Our approach makes use of a sieve bootstrap scheme based on residual resampling from autoregressive approximations the order of which increases with the sample size at a suitable rate. The first-order asymptotic correctness of the sieve bootstrap for testing the stationarity hypothesis is established for a subclass of linear processes. The small-sample properties of the method are also investigated by means of Monte Carlo experiments. [Copyright &y& Elsevier]

In this paper, a new estimation method is introduced for the quantile spectrum, which uses a parametric form of the autoregressive (AR) spectrum coupled with nonparametric smoothing. The method begins with quantile periodograms which are constructed by trigonometric quantile regression at different quantile levels, to represent the serial dependence of time series at various quantiles. At each quantile level, we approximate the quantile spectrum by a function in the form of an ordinary AR spectrum. In this model, we first compute what we call the quantile autocovariance function (QACF) by the inverse Fourier transformation of the quantile periodogram at each quantile level. Then, we solve the Yule–Walker equations formed by the QACF to obtain the quantile partial autocorrelation function (QPACF) and the scale parameter. Finally, we smooth QPACF and the scale parameter across the quantile levels using a nonparametric smoother, convert the smoothed QPACF to AR coefficients, and obtain the AR spectral density function. Numerical results show that the proposed method outperforms other conventional smoothing techniques. We take advantage of the two-dimensional property of the estimators and train a convolutional neural network (CNN) to classify smoothed quantile periodogram of earthquake data and achieve a higher accuracy than a similar classifier using ordinary periodograms. [ABSTRACT FROM AUTHOR]