Research Catalog

Title

Neural, novel & hybrid algorithms for time series prediction / Timothy Masters.

Author

Masters, Timothy

Publication

New York : John Wiley & Sons, [1995], ©1995.

Items in the Library & Off-site

Holdings

Details

Alternative Title

Neural, novel, and hybrid algorithms for time series prediction

Subject

• Neural networks (Computer science)
• Algorithms

Bibliography (note)

• Includes bibliographical references (p. [489]-507) and index.

Contents

• 1. Preprocessing. Things Everyone Should Know. Centering and Detrending. Data Reduction and Orthogonalization. Filtering for Information Selection -- 2. Subduing Seasonal Components. The Problem of Harmonics. A Seasonality Example -- 3. Frequency-Domain Techniques I: Introduction. Introduction to the Frequency Domain. The Discrete Fourier Transform. The Power Spectrum. The Maximum Entropy Power Spectrum. Power Spectrum Examples and Summary -- 4. Frequency-Domain Techniques II: Filters and Features. What Is a Digital Filter? Filtering in the Frequency Domain. Bandpass Filters. Lowpass and Highpass Filters. Implementation Details and Sample Code. In-quadrature Filters. Quadrature-Mirror Filters -- 5. Wavelet and QMF Features. Feature Presence versus Feature State. The Width Dilemma. Wavelet Features. The Morlet Wavelet -- 6. Box-Jenkins ARMA Models. Overview of the ARMA Paradigm. Multivariate ARMA Models. Training the Multivariate ARMA Model.
• Designing and Testing ARMA Models. Computing Lagged Correlations. A Multivariate Example -- 7. Differencing. Stationarity. ARIMA Models. Seasonal Differencing. An Example of Differencing -- 8. Robust Confidence Intervals. Overview. Sampling the Prediction Errors. Compensating for Differencing and Transformations. From Errors to Confidence Intervals. Confidence in the Confidence. Multiplicative Confidence Intervals. Confidence Intervals in Action -- 9. Numerical and Statistical Tools. Random Numbers. Singular Value Decomposition. Deterministic Optimization. Stochastic Optimization. Eigenvalues and Eigenvectors. Fourier Transforms. Data Reduction and Orthogonalization. Autoregression by Burg's Algorithm -- 10. Neural Network Tools. Training and Test Sets. Generic Network Parameters. Generic Learning Parameters. Generic Neural Networks. Multiple-Layer Feedforward Networks. Probabilistic Neural Networks -- 11. Using the NPREDICT Program. Using This Manual.
• General Commands. Working with Signals. The Power Spectrum and its Relatives. Data Reduction and Orthogonalization. Filters. Autocorrelation and Related Operations. Box-Jenkins ARMA/ARIMA Prediction. Neural Network Training and Test Sets. Neural Network Models. Neural Network Prediction. Alphabetical Glossary of Commands. Validation Suite. Appendix. Disclaimer. Disk Contents. Hardware and Software Requirements. Compiling and Linking the NPREDICT Programs. Data Files.

ISBN

0471130419 (paper : alk. paper)

LCCN

95031203

OCLC

• 32853567
• ocm32853567

Owning Institutions

Columbia University Libraries