On Optimality of Bayesian Wavelet Estimators

Abstract.  We investigate the asymptotic optimality of several Bayesian wavelet estimators, namely, posterior mean, posterior median and Bayes Factor, where the prior imposed on wavelet coefficients is a mixture of a mass function at zero and a Gaussian density. We show that in terms of the mean squared error, for the properly chosen hyperparameters of the prior, all the three resulting Bayesian wavelet estimators achieve optimal minimax rates within any prescribed Besov space     for p  ≥ 2. For 1 ≤  p  < 2, the Bayes Factor is still optimal for (2 s +2)/(2 s +1) ≤  p  < 2 and always outperforms the posterior mean and the posterior median that can achieve only the best possible rates for linear estimators in this case.     

Real nonparametric regression using complex wavelets

Summary.  Wavelet shrinkage is an effective nonparametric regression technique, especially when the underlying curve has irregular features such as spikes or discontinuities. The basic idea is simple: take the discrete wavelet transform of data consisting of a signal corrupted by noise; shrink or remove the wavelet coefficients to remove the noise; then invert the discrete wavelet transform to form an estimate of the true underlying curve. Various researchers have proposed increasingly sophisticated methods of doing this by using real-valued wavelets. Complex-valued wavelets exist but are rarely used. We propose two new complex-valued wavelet shrinkage techniques: one based on multiwavelet style shrinkage and the other using Bayesian methods. Extensive simulations show that our methods almost always give significantly more accurate estimates than methods based on real-valued wavelets. Further, our multiwavelet style shrinkage method is both simpler and dramatically faster than its competitors. To understand the excellent performance of this method we present a new risk bound on its hard thresholded coefficients.     

Two preprocessing algorithms for climate time series

We propose two preprocessing algorithms suitable for climate time series. The first algorithm detects outliers based on an autoregressive cost update mechanism. The second one is based on the wavelet transform, a method from pattern recognition. In order to benchmark the algorithms'' performance we compare them to existing methods based on a synthetic data set. Eventually, for exemplary purposes, the proposed methods are applied to a data set of high-frequent temperature measurements from Novi Sad, Serbia. The results show that both methods together form a powerful tool for signal preprocessing: In case of solitary outliers the autoregressive cost update mechanism prevails, whereas the wavelet-based mechanism is the method of choice in the presence of multiple consecutive outliers.     

Uniform Convergence of Compactly Supported Wavelet Expansions of Gaussian Random Processes

New results on uniform convergence in probability for expansions of Gaussian random processes using compactly supported wavelets are given. The main result is valid for general classes of non stationary processes. An application of the obtained results to stationary processes is also presented. It is shown that the convergence rate of the expansions is exponential.     

基于LSTM神经网络的金融时间序列预测

本文提出将小波分析与纳入时间序列依赖特征的长短期记忆（LSTM）神经网络相结合，构建金融时间序列数据预测模型，以克服现有模型对金融时间序列数据非平稳、非线性、序列相关等复杂特征以及数据间非线性交互关系无法反映的缺陷。同时，以道琼斯工业指数日收盘价为例，探究LSTM神经网络对实际金融时间序列数据的预测能力，比较其与多层感知机、支持向量机、K近邻、GARCH四种模型的预测效果。实证结果表明LSTM神经网络具有更高的预测精度，能够有效预测金融时间序列数据的长短期动态变化趋势，说明了其对金融时间序列数据预测的适用性与有效性。此外，对金融时间序列数据进行小波分解与重构，可有效提高LSTM预测模型的泛化能力，以及对长短期动态趋势的预测精度。     

违约预测的小波结构模型研究

本文针对上市公司违约预测问题,按照行业类型对我国2009年的上市企业进行分层抽样,构建了小波结构模型.小波结构模型通过应用小波变换来分解上市公司日收益序列,进而对低频序列和高频序列分别构建预测模型,再依据预测模型对未来收益进行预测,最后使用小波逆变换重构预测收益序列.通过小波结构模型可以避免时间序列模型进行收益波动预测的累加计算过程.在结合我国上市公司的实际数据对这两种模型的校验中,可以发现小波结构模型比时序结构模型在违约预测上有更好的识别力和准确度.     

Analyzing time–frequency relationship between oil price and exchange rate in Pakistan through wavelets

This study analyzed the time–frequency relationship between oil price and exchange rate for Pakistan by using measures of continuous wavelet such as wavelet power, cross-wavelet power, and cross-wavelet coherency (WTC). The results of cross-wavelet analysis indicated that covariance between oil price and exchange rate is unable to give clear-cut results, but both variables have been in phase and out phase (i.e. they are anti-cyclical and cyclical in nature) in some or other durations. However, results of squared wavelet coherence disclose that both variables are out of phase and real exchange rate was leading during the entire period studied, corresponding to the 10–15 months’ scale. These results are the unique contribution of the present study, which would have not been drawn if one would have utilized any other time series or frequency domain-based approach. This finding provides evidence of anti-cyclical relationship between oil price and real effective exchange rate; however, in most of the period studied, real exchange rate was leading and passing anti-cycle effects on oil price shocks which is the major contribution of the study.     

Forecasting interest rates: a comparative assessment of some second-generation nonlinear models

Modeling and forecasting of interest rates has traditionally proceeded in the framework of linear stationary methods such as ARMA and VAR, but only with moderate success. We examine here three methods, which account for several specific features of the real world asset prices such as nonstationarity and nonlinearity. Our three candidate methods are based, respectively, on a combined wavelet artificial neural network (WANN) analysis, a mixed spectrum (MS) analysis and nonlinear ARMA models with Fourier coefficients (FNLARMA). These models are applied to weekly data on interest rates in India and their forecasting performance is evaluated vis-à-vis three GARCH models [GARCH (1,1), GARCH-M (1,1) and EGARCH (1,1)] as well as the random walk model. Both the WANN and MS methods show marked improvement over other benchmark models, and may thus hold out several potentials for real world modeling and forecasting of financial data.     

基于小波变系数模型的CPI权重估计

居民消费价格指数（CPI）是反映宏观经济运行的重要指标之一,因此其编制中所采用权重的取值和调整特征一直是各界所关心的问题,但目前中国CPI编制中所采用的权重大小一直没有权威机构的正式公布。根据2000—2010年间的CPI数据,采用基于小波的变系数模型对权重进行了估计。估计结果揭示了中国近10年中CPI权重的动态变化特征,这种变化特征也反映了近10年中居民消费结构的动态调整。