Brain Cancer Antibody Display Classification

This article explores real data on brain cancer. This type of biological data has a few particularities like a great number of attributes – antibodies and genes. However the number of entries is rather small because the data have to be obtained from real patients. This process is time consuming and very costly. Due to that, this research provides detailed data description as well as analyzes their particularities, type and structure. Correspondingly, classification rules are also difficult to discover. This research is dedicated to finding applications of classification methods aimed at determining interconnections that could be used to classify brain cancer. Working exactly with such unique data has a great practical value, because the data obtained can be used in future to continue the research and in practical diagnostics with the possibility to offer the data to biologists for interpretation. To speed up the obtaining of interconnections, only important attributes were used. Various methods of interconnection determination were employed. Conclusions about this type of data analysis, obtaining classification rules and the precision of obtained rules are made and directions of future work are outlined.