: 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.

eMedicine创建于1996年，由近万名临床医师作为作者或编辑参与此临床医学知识库的建设，其中编辑均是来自美国哈佛、耶鲁、斯坦福、芝加哥、德克萨斯、加州大学等各分校医学院的教授或副教授。

eMedicine

This chapter introduces the subject of statistical pattern recognition (SPR). It starts by considering how features are defined and emphasizes that the nearest neighbor algorithm achieves error rates comparable with those of an ideal Bayes’ classifier. The concepts of an optimal number of features, representativeness of the training data, and the need to avoid overfitting to the training data are stressed. The chapter shows that methods such as the support vector machine and artificial neural networks are subject to these same training limitations, although each has its advantages. For neural networks, the multilayer perceptron architecture and back-propagation algorithm are described. The chapter distinguishes between supervised and unsupervised learning, demonstrating the advantages of the latter and showing how methods such as clustering and principal components analysis fit into the SPR framework. The chapter also defines the receiver operating characteristic, which allows an optimum balance between false positives and false negatives to be achieved.

Statistical Pattern Recognition

The detection of recorded epileptic seizure activity in EEG segments is crucial for the localization and classification of epileptic seizures. However, since seizure evolution is typically a dynamic and nonstationary process and the signals are composed of multiple frequencies, visual and conventional frequency-based methods have limited application. In this paper, we demonstrate the suitability of the time-frequency ( t-f) analysis to classify EEG segments for epileptic seizures, and we compare several methods for t- f analysis of EEGs. Short-time Fourier transform and several t-f distributions are used to calculate the power spectrum density (PSD) of each segment. The analysis is performed in three stages: 1) t-f analysis and calculation of the PSD of each EEG segment; 2) feature extraction, measuring the signal segment fractional energy on specific t-f windows; and 3) classification of the EEG segment (existence of epileptic seizure or not), using artificial neural networks. The methods are evaluated using three classification problems obtained from a benchmark EEG dataset, and qualitative and quantitative results are presented.

Epileptic Seizure Detection in EEGs Using Time–Frequency Analysis

Time--frequency feature representation using energy concentration: An overview of recent advances

https://espace.library.uq.edu.au/view/UQ:7736/SIGPRO-3123.pdf

IF estimation for multicomponent signals using image processing techniques in the time-frequency domain

The nonstationary and multicomponent nature of newborn EEG seizures tends to increase the complexity of the seizure detection problem. In dealing with this type of problems, time-frequency-based techniques were shown to outperform classical techniques. This paper presents a new time-frequency-based EEG seizure detection technique. The technique uses an estimate of the distribution function of the singular vectors associated with the time-frequency distribution of an EEG epoch to characterise the patterns embedded in the signal. The estimated distribution functions related to seizure and nonseizure epochs were used to train a neural network to discriminate between seizure and nonseizure patterns.

Time-Frequency Feature Extraction of Newborne EEG Seizure using SVD-Based Techniques

Time-frequency peak filtering (TFPF) allows the reconstruction of signals from observations corrupted by additive noise by encoding the noisy signal as the instantaneous frequency (IF) of a frequency modulated (FM) analytic signal. IF estimation is then performed on the analytic signal using the peak of a time-frequency distribution (TFD) to recover the filtered signal. This method is biased when the peak of the Wigner-Ville distribution (WVD) is used to estimate the encoded signal's instantaneous frequency. We characterize a class of signals for which the method implemented using the pseudo WVD is approximately unbiased. This class contains deterministic bandlimited nonstationary multicomponent signals in additive white Gaussian noise (WGN). We then derive the pseudo WVD window length that gives a reduced bias when TFPF is used for signals from this class. Testing of the method on both synthetic and real life newborn electroencephalogram (EEG) signals shows clean recovery of the signals in noise level down to a signal-to-noise ratio (SNR) of -9 dB.

Signal enhancement by time-frequency peak filtering

The detection of seizure in the newborn is a critical aspect of neurological research. Current automatic detection techniques are difficult to assess due to the problems associated with acquiring and labelling newborn electroencephalogram (EEG) data. A realistic model for newborn EEG would allow confident development, assessment and comparison of these detection techniques. This paper presents a model for newborn EEG that accounts for its self-similar and nonstationary nature. The model consists of background and seizure submodels. The newborn EEG background model is based on the short-time power spectrum with a time-varying power law. The relationship between the fractal dimension and the power law of a power spectrum is utilized for accurate estimation of the short-time power law exponent. The newborn EEG seizure model is based on a well-known time-frequency signal model. This model addresses all significant time-frequency characteristics of newborn EEG seizure which include; multiple components or harmonics, piecewise linear instantaneous frequency laws and harmonic amplitude modulation. Estimates of the parameters of both models are shown to be random and are modelled using the data from a total of 500 background epochs and 204 seizure epochs. The newborn EEG background and seizure models are validated against real newborn EEG data using the correlation coefficient. The results show that the output of the proposed models have a higher correlation with real newborn EEG than currently accepted models (a 10% and 38% improvement for background and seizure models, respectively)

/pdf/a-nonstationary-model-of-newborn-eeg-3r3nxnjuii.pdf

A Nonstationary Model of Newborn EEG

/pdf/time-frequency-feature-extraction-of-newborn-eeg-seizure-3qte8uujdh.pdf

Mostefa Mesbah

Papers

IF estimation for multicomponent signals using image processing techniques in the time-frequency domain

Time-Frequency Feature Extraction of Newborne EEG Seizure using SVD-Based Techniques

Signal enhancement by time-frequency peak filtering

A Nonstationary Model of Newborn EEG

Time-frequency feature extraction of newborn EEG seizure using SVD-based techniques