: This thesis applies neural network feature selection techniques to multivariate time series data to improve prediction of a target time series. Two approaches to feature selection are used. First, a subset enumeration method is used to determine which financial indicators are most useful for aiding in prediction of the S&P 500 futures daily price. The candidate indicators evaluated include RSI, Stochastics and several moving averages. Results indicate that the Stochastics and RSI indicators result in better prediction results than the moving averages. The second approach to feature selection is calculation of individual saliency metrics. A new decision boundary-based individual saliency metric, and a classifier independent saliency metric are developed and tested. Ruck's saliency metric, the decision boundary based saliency metric, and the classifier independent saliency metric are compared for a data set consisting of the RSI and Stochastics indicators as well as delayed closing price values. The decision based metric and the Ruck metric results are similar, but the classifier independent metric agrees with neither of the other metrics. The nine most salient features, determined by the decision boundary based metric, are used to train a neural network and the results are presented and compared to other published results. (AN)

/pdf/nonlinear-time-series-analysis-1j3a35n4y3.pdf

Nonlinear Time Series Analysis.

Introduction to quality engineering : designing quality into products and processes

Complex network approaches to nonlinear time series analysis

Dynamics of Structures: Theory and Applications to Earthquake Engineering by Anil K. Chopra

Output-only modal identification plays an important role in the structural health monitoring of large-scale structures. In recent years, blind source separation (BSS) has achieved great suc...

Clustering Number Determination for Sparse Component Analysis during Output-Only Modal Identification

Selection of optimal threshold to construct recurrence plot for structural operational vibration measurements

Practical ambient excitations of engineering structures usually do not comply with the stationary-white-noise assumption in traditional operational modal analysis methods due to heavy traffic, wind guests, and other disturbances. In order to eliminate spurious modes induced by non-white noise inputs, the improved stochastic subspace identification based on a delay index is proposed in this paper for a representative kind of stationary non-white noise ambient excitations, which have nonzero autocorrelation values near the vertical axis. It relaxes the stationary-white-noise assumption of inputs by avoiding
corresponding unqualified elements in the Hankel matrix. Details of the improved stochastic subspace identification algorithms and determination of the delay index are discussed. Numerical simulations on a four-story frame and laboratory vibration experiments on a simply supported beam have demonstrated the accuracy and reliability of the proposed method in eliminating spurious modes under non-white noise ambient excitations.

Operational modal analysis of structures by stochastic subspace identification with a delay index

Non-stationary assessment of structural operational measurements using recurrence quantification analysis

A method based on Hilbert Transform is presented to separate components or to extract the envelope from the non-stationary amplitude modulated signals in the product model where the non-stationary signal is the product of the non-stationary slowly-varying component and stationary fast component. Numerical examples are given to illustrate the efficiency of the presented method. It is demonstrated that the stationary part can be successfully separated from the non-stationary amplitude modulated signals in the product model.

Yi-Ding Hu

Papers

Selection of optimal threshold to construct recurrence plot for structural operational vibration measurements

Operational modal analysis of structures by stochastic subspace identification with a delay index

Non-stationary assessment of structural operational measurements using recurrence quantification analysis

Separating the components of amplitude modulated non-stationary signal based on Hilbert transform