A data classification technique that provides qualitative and quantitative information inspired by the chromatographic separation method of substances

• Published in: Proceedings ... International Conference on Soft Computing & Machine Intelligence ISCMI ... (Online) pp. 106 - 111
• Main Author: Swiecicki, Mariusz
• Format: Conference Proceeding
• Language: English
• Published: IEEE 22.11.2024
• Subjects: chromatographic separation; classification algorithm; Classification algorithms; data mining; Decision making; Distributed databases; Gas chromatography; Landsat; Machine intelligence; natural computing algorithms; Satellites; Thyroid; Vectors
• ISSN: 2640-0146
• DOI: 10.1109/ISCMI63661.2024.10851624

Currently, there are several significant limitations in the issues related to the exploration of large data sets using classical classification methods, the results of which are used in the decision-making process. The first significant limitation in the use of classical classification methods is the need to ensure a constant size of the data. The second type of limitation is related to the dimension of the data. The last type of limitation that occurs when using classical classification algorithms is related to the situation that a given input vector may contain data belonging to many classes simultaneously. On the other hand, as a result of processing large data sets, we want to obtain not only qualitative information, but also quantitative information, which is equally important in the decision-making process. It seems that one of the methods of solving the above problems in data classification is the adaptation of classification mechanisms used, for example, in analytical chemistry to identify complex chemical compounds in the tested mixture. The article proposes an innovative algorithm for the classification of multidimensional data based on the chromatographic separation method. The article presents a distributed data classification algorithm based on the gas chromatography technique, which in issues related to the classification of large data sets is not subject to the above limitations and provides quantitative and qualitative information. The article also presents the classification results of the algorithm for selected - standard data sets. This work shows that based on the proposed chromatographic separation method, we also obtain quantitative information, not only qualitative information, which is a significant advantage in comparison to classical classification methods.