In a computerized method for classification of a time series containing a prescribable number of samples, such as an electrical signal, a parameterized, dynamic set is determined in the computer from the samples in the time series, the set identifying non-linear correlations between the samples of the time series. A classification of the time series is implemented in the computer on the basis of the parameterized, dynamic multiplicity.

The invention detects signs of emerging illness, such as flu, skin cancer, backache, etc., among the general population and individuals. The detection of symptoms of illness is performed through the utilization of embedded devices equipped with various sensors, such as cameras, glasses, wrist watches, TVs, fire warning systems, and having the ability to analyze the detected information and to transmit that information via wireless and regular communication channels to a central server for a more detailed analysis and possible action. The information about locally detected symptoms is gathered at central location and processed to ascertain whether there is a new epidemic of a flu, therefore enabling early shipment of a flu vaccine which prevents the spread of the disease.

A method and apparatus for improved modeling of random or stochastic systems such as financial markets and instruments. The Black-Scholes model for the price of a risky asset is replaced by an improved model which uses a fractal activity time process instead of a conventional time parameter. The new model is incorporated into an algorithm for analyzing empirical data in order to determine whether an observed random process exhibits multifractal or single fractal behavior. If the algorithm finds multifractal or single fractal behavior, the new model can be used, thereby providing improved accuracy over the Black-Scholes model. The algorithm can be implemented as an interactive procedure using a standard computer hardware and software platform operating under the control of suitable software for implementing the computing steps of the disclosed algorithm.

A time series that is established by a measured signal of a dynamic system, for example a quotation curve on the stock market, is modelled according to its probability density in order to be able to make a prediction of future values. A non-linear Markov process of the order m is suited for describing the conditioned probability densities. A neural network is trained according to the probabilities of the Markov process using the maximum likelihood principle, which is a training rule for maximizing the product of probabilities. The neural network predicts a value in the future for a prescribable number of values m from the past of the signal to be predicted. A number of steps in the future can be predicted by iteration. The order m of the non-linear Markov process, which corresponds to the number of values from the past that are important in the modelling of the conditioned probability densities, serves as parameter for improving the probability of the prediction.

Apparatus for graphically representing a substantially periodic signal having substantially periodic segments, comprising a graphic transformer for converting the periodic segments into encodings respectively for arrangement in a first dimension and successively indexing the encodings to form a two-dimensional coded image, thereby to represent, within the image, dynamic changes of the periodic segments over the signal.