The automated machine learning solution covers a wide range of spectroscopic initial data preparation and analysis (regression, classification, quantification) methods and their parameter selection algorithms. The solution is intended for both science and industry to deal with chemometric tasks and needs to calibrate or develop spectral classification or interpretation algorithms.
Projects related to the analysis of spectroscopic data often encounter the fact that the software on the market is expensive, complex / requiring relevant knowledge, limited functionality, and not automated. This affects the cost of modeling, the need for human resources, the efficiency of the project and the average efficiency.
Intelligent system functions allow users to automate the selection of the optimal sequence of methods and their parameters. The system is based on the architecture of evolutionary algorithms, which allows to shorten the search time and ensure that the resulting algorithm configuration is close to optimal. In addition, the system uses unique methods for the detection and selection of spectral markers.
An automated machine learning solution is a suite of machine learning products that enables users with limited knowledge of computer learning or chemometrics to create high-quality models to meet their scientific or business needs.