Chemometrics

 

Chemometrics is applied in cases where the direct measurement of properties, such as concentration, of substances and systems is complicated. The aim is to model linear and non-linear relationships of complex data, to classify their objects, identify their significant characteristics as well as to visualize these data.

Methods

  • Modeling using statistical methods (PCA, MLR, PCR, PLS, QPLS)
  • Modeling and classification using Neuronal Networks
  • Classification using SOM, LVQ, SIMCA, LDA
  • Selection of variables using genetic algorithms, Pruning algorithms and Greedy algorithms
  • Optimisation via design of experiments

Applications

  • Quantification of multi analyte mixtures using time-resolved sensor measurements in the gaseous phase
  • Analysis of data acquired by biosensors using Neuronal Networks
  • Identification of trends based on highly variable data sets (see project AQUATERRA)

Literature

  • Different approaches to multivariate calibration of nonlinear sensor data, F. Dieterle, S. Busche, G. Gauglitz, Analytical and Bioanalytical Chemistry (2004), 380(3), 383-396
  • Growing neural networks for a multivariate calibration and variable selection of time-resolved measurements, F. Dieterle, S. Busche, G. Gauglitz, Analytica Chimica Acta (2003), 490(1-2), 71-83
  • Genetic algorithms and neural networks for the quantitative analysis of ternary mixtures using surface plasmon resonance F. Dieterle, B. Kieser, G. Gauglitz, Chemometr Intell Lab 65, 67-81 (2003)
  • Multi-analyte assay for triazines using cross-reactive antibodies and neural net-works, S. Reder, F. Dieterle, H. Jansen, S. Alcock, G. Gauglitz, Biosens Bioelectron 19, 447-455 (2003)
  • Urinary Nucleosides as Potential Tumor Markers Evaluated by Learning Vecto Quantization, F. Dieterle, S. Müller-Hagedorn, H. Liebich, G. Gauglitz, Artif Intell Med 28, 265-279 (2003)