Fully automatic algorithm for analysis of gas chromatography data
Abstract: The use of chromatography is widely spread today, ranging from industry and medicine to investigatory police work. However the procedures for analyzing the output data have been developed since the launch of the technique in the beginning of the 20th century. From cutting in paper to advanced computerized algorithms. Today there are many commercially available analysis programs adapted for chromatography. However they are often "black boxes" and much research can therefore be found on how to improve the chromatographic work. Although most of the results of the methods found in literature are acceptable, it is frequently one part in the process that is not fully automated. Some require input parameters from the user, which introduces subjectivity and in addition user- experience dependent results. In this work an entirely automatic algorithm is developed where the only input is the .txt file produced by the chromatograph. On an average PC the time for one scan is under 10s for a dataset of 5000 points. This algorithm can be applied to a large amount of similar scans in order to build up a large database for later machine learning.
AT THIS PAGE YOU CAN DOWNLOAD THE WHOLE ESSAY. (follow the link to the next page)