|
When analysing data, we employ a wide range of classic and modern analysis methods. We use established procedures as well as the latest developments in the fields of statistics and machine learning. We refine and adapt existing methods as needed, optimising them for the requirements presented by our clients.
Below you will find a selection of our methods:
General statistics
- Univariate and multivariate analytical techniques
- Parametric and non-parametric density estimates
- Testing procedures and error calculation, definition of confidence intervals
- Dimensional models (factor and cluster analyses)
Machine learning
- Detection of irregular events (anomaly detection)
- Definition and calculation of scores
- Prediction and analysis of (non-)linear time series, clustering, classification and visualisation
- Broad spectrum of neural networks, ranging from decision trees to support vector machines and Gaussian processes
Visualisation and embedding
- Linear and non-linear dimensional reductions
- Graphical depiction of complex correlations
- Linear and non-linear source separation methods
|
