Peter Wentzell

FoS Killam Professor (2014)
Chemistry Department
  

Education:

Mining Chemical Measure-ments

Measurements on chemical systems span the range of the very simple, such as pH assessment with a test strip, to the very complex, including multidimensional separations, hyphenated instrumentation, and hyperspectral imaging.  More and more, difficult questions in areas such as medicine, the environment, industry, pharmaceutical discovery and forensics require multivariate analytical measurements.  It is the goal of chemometrics to extract the desired information from such chemical measurements by applying advanced statistical, mathematical and computer based strategies.

The objective of research in Dr. Wentzell's group is to develop new and better chemometric tools to extract information from multivariate measurements.  Areas of research include exploratory data analysis, pattern recognition (clustering, classification), multivariate calibration, signal processing and curve resolution.  Particular emphasis is placed on the characterization of measurement noise and its effect on multivariate analysis methods.  Optimal methods for data analysis are developed through maximum likelihood implementations of widely used methods such as principal components analysis.  Areas of application range from biology (DNA microarrays, metabolomics, proteomics) to the environment (source-receptor modeling) and food science.