Data processing, forecasting, numerical methods

PNN Solutions is R&D department of PNN company that develops innovative data mining and real time signal processing methods and algorithms and is responsible for their applications in computational chemistry, neurophysiology, medicine (ECG & EEG analysis), financial market analysis and forecasting and other fields.
PNN research team includes the experts in the fields of modeling and forecasting on experimental data, decision support and pattern recognition, linear and nonlinear analyses, signal processing and numerical methods.

PNN provides both scientific solution and software implementation.
	The main algorithms families PNN Solutions deals with:
	Polynomial Neural Network (PNN) algorithms    >>>
	Phase Space Technique (PST) algorithms
	Epilepsy seizure prediction algorithm >>>

Polynomial Neural Network (PNN) algorithm - GMDH-type neural network - one of the most promising methods developed by PNN Research team. The goal of the algorithm application is to extract knowledge from experimental data and to determine its best mathematical description. PNN - a self-organizing multi-layered iterative algorithm that automatically provides linear and non-linear polynomial regression models. The PNN embodies the advantages of Multiple Linear Regression (MLR) and Artificial Neural Networks (ANNs) into a single entity. It can model both linear and non-linear relationships like ANNs, and it yields a polynomial regression equation like MLR for easy interpretation. This algorithm provides robust results in the presence of correlated and irrelative variables or/and outliers. The results of this algorithm can be easily interpreted. More >>>

Phase Space Technique (PST) is another innovative method for nonlinear analysis, signal processing and pattern recognition developed by PNN Research. This approach considers signals in phase space and accounts for two different types of noise: additive and perturbative. The first type, additive noise, contributes to distortion of the absolute values of signal peaks. The second type, perturbative noise, contributes to variations of the retention times of signal peaks and distorts the time scale. The ability to consider both types of noise significantly distinguishes it from existing methods of data analysis, which are usually designed to treat only the additive noise. Analysis of signals in phase space eliminates the problem of perturbation noise and enables detection and comparison of similar signal segments realized at different retention times. PNN Company PST algorithms implementation not only gives the unique opportunity for signal processing but also provides the efficient real time software.

 
Scientific research was performed in collaboration with

• GenChemiCs, Princeton, NJ, US
• University of Lausanne, Switzerland
• University of Missouri-St. Louis, US
• Center for Molecular Design, University of Portsmouth, UK

• Nelson&Moritz, Hamburg, Germany