Innovative algorithm of statistical data analysis method for optical spectrometer

Author: george kachlishvili
Keywords: innovative spectroscope, data analysis, machine learning
Annotation:

Data analysis using machine learning methodologies (including possibly artificial intelligence algorithms) is a relatively new approach based on statistical resampling. All this is connected, on the one hand, with advanced digital computing technologies, and, on the other hand, with a total transition from analog devices to digital ones. The resources of modern processors make it possible to reduce the sampling parameter in an analog signal so that the information loss tends to zero (the Shannon-Naquist theorem) and, among other things, repeatedly receive data from the object under study, possibly from different space-time coordinates. From the course of molecular and atomic physics, it is known that spectroscopic data contain huge, but rather delicate information, which is of great importance in different areas, and correct and fast decoding of the encoded information is a rather serious problem. And also with the development of the technology of optical spectrometers, there was a multiple increase in resolution; although resolution can be improved not only with improved spectrometric technologies, but also with the use of dedicated data analysis. It should be noted that the data obtained from a high-resolution optical spectrometer encoded a huge amount of useful information that is almost impossible to read with the naked eye.