Pattern Analysis - Development of Customized Algorithmic Solutions

Ref-Nr:

Technology abstract

The german company offers algorithmic solutions for the analysis of complex systems and processes. Experienced experts (with strong background in natural sciences) develop solution strategies in cooperation with their clients, leading to customized solutions for tasks, where standard techniques are doomed to fail. Thereto the company  provides proprietary algorithms and state-of-the-art data analysis techniques from the field of data mining .

Technology Description

The german company offers consulting and customized algorithmic solutions for complex data analysis problems. The methods, primarily developed for the scientific analysis of complex systems, have been successfully applied to many real world problems.
The company has expert knowledge in (scientific) data analysis in a wide field of applications and is experienced in the successful transfer to its clients.
A special feature of the partially proprietary methods is the use of measures derived from information theory, designed for the investigation of nonlinear dynamical systems. Combined with elements from graph theory, these methods are especially suited for data mining problems. The portfolio is completed by methods from the field of machine learning such as neural networks. These broad methodological skills enable innovative solutions for challenging data analysis problems and various data modalities.
Data analysis tasks such as classification, segmentation, and prediction can be solved very frequently by pattern recognition in high-dimensional (artificial) state-spaces. Therefore, concepts and strategies have been established that allow for:

  • Assembling and homogenizing data sets with respect to their contextual information content
  • Dimensionality reduction of state-spaces by means of redundancy analyses,
  • Reliability assessment of findings with the help of surrogate data
  • Recognition and reduction of noise, artefacts and measurement errors (data cleaning).

Due to their general formulation most of these methods are not restricted to a specific data modality. Besides image processing CSYSC has expert knowledge of many years in the analysis of time series and spectral data. The applications range from biosignal processing in medicine (ECG, cardiology; EEG, neurology), innovative evaluation of audio data, analysis of spectra (chemometry, nutritional components) to the investigation of time series in the energy sector (concerning trade, consumption, PV power generation).

Innovations & Advantages

  • Proprietary algorithms and strategies based e.g. on nonlinear dynamics, information theory (scaling behavior, etc.)
  • Comprehensive pre-processing of the data (data cleaning, consistency checks …)
  • Customized solutions for specific demands
  • Retrieval of information yet hidden in the data
  • Solutions for high data throughput employing expert knowledge in solutions (big data)

Statistical evaluation (descriptive and explorative)

Further Information

Current and Potential Domains of Application

The applications range from biosignal processing in medicine (ECG, cardiology; EEG, neurology), innovative evaluation of audio data, analysis of spectra (chemometry, nutritional components) to the investigation of time series in the energy sector (concerning trade, consumption, PV power generation).