Contact us by email if you have any questions or wish to engage us in some sort of consulting arrangement related to data analysis.

consult (at) chemometrics-analysis.com

Example projects / suggestions:

WOULD YOU LIKE TO CREATE CLASSIFICATION / PREDICTION MODELS FROM YOUR DATA THAT CAN BE ENCODED INTO SIMPLE MODULES THAT ANYONE CAN USE?

WOULD YOU LIKE MODELS DEVELOPED FROM SUCH ALGORITHMS AS:
PRINCIPAL COMPONENT ANALYSIS
PARTIAL LEAST SQUARES
PROBABILISTIC NEURAL NETWORKS
BAYESIAN BELIEF NETWORKS
SUPPORT VECTOR MACHINES
BACK PROPAGATION NEURAL NETWORKS

many of these algorithms can be designed to include variable subset selection using genetic analysis, denoising, peak alignment, and vatious scaling / normalization functions...

DO YOU HAVE A LOT OF DATA THAT NEED TO BE ORGANIZED?

DO YOU NEED TO ASSEMBLE A DATABASE FROM WHICH INFORMATION CAN BE READILY EXTRACTED, VISUALIZED AND ANALYZED?





DEVELOPMENT OF A DATABASE AND RELATED TOOLS FOR CHEMOMETRICS DATA ANALYSIS

The following material demonstrates some of the real-life programming possibilities that we have experience in ... the exact form of the interface can be determined / designed by the user



objectives are to provide the power of a database with extensive functionality and a custom-designed interface that minimizes the need for time-consuming new user training while still allowing complex queries and visualization to be achieved

we provide the additional, customized functionality as demosntrated from the figures below



NOTE THAT VARIOUS SIGNIFICANT PREPROCESSING OPTIONS CAN BE BUILT INTO THIS MODULE (PEAK ALIGNMENT; DENOISING; VARIOUS PREPROCESSING OPTIONS)

MOST PREPROCESSING OPTIONS CAN BE AUTOMATED INTO A SINGLE CLICK OF A BUTTON





various classification / prediction algorithms can be applied to your problem(s)

THE MODELS THEN DEVELOPED CAN BE ENCODED IN MODULES THAT CAN THEN BE USED BY USERS WITH MINIMAL EXPERIENCE.
NO ADDITIONAL OR THIRD-PARTY SOFTWARE IS REQUIRED FOR USE.





example of an output display that demonstrates that multiple algorithms can be applied to multiple analytical measurements with a single consensus classification or prediction made





the database code also includes an extensive plot utility for publication-quality graphics



we can define any sort of interface according to user requirements or multiple interface options can be designed for users with different levels of experience



users can access any part of the database and visualize the data; statistical analyses can be applied to the chosen data



below we demonstrate how various GC variables in a database can be individually accessed and plotted



the resulting plots can be annotated as desired