Practical chaos time series analysis with financial applications

We describe the practical implementation of the nonlinear (chaos) time series analysis based on the paradigm of deterministic chaos. Some important techniques of statistical test for nonlinearity, phase space reconstruction, and nonlinear prediction are discussed with some applications to finance. The use of the nonlinear time series analysis is illustrated with particular emphasis on issues of choices of time delay coordinate and embedding parameters. Finally, we propose the weighted embedding method which is found to work well in financial applications