Evaluation of Bacterial Biofilm Category Change Due to the Use of Different Signaling Molecules Using Random Forest Classifier

Bacteria have a unique ability to form groups, known as biofilm, which provides nutrients and protection for their inhabitants. Based on its structure and density, i.e. the number of bacteria forming it, biofilm can be classified into four categories. Its classification can be performed by applying two types of tests: qualitative, and quantitative, with the latter one being more reliable. However, laboratory analyses turned out to be time and money-consuming. This is where machine learning came to use. Random Forest is a supervised, ensemble machine learning algorithm that proved itself to be a powerful tool in medical analyses, and we used it in this research for the evaluation of bacterial biofilm category change due to the use of different signaling molecules, but prior to that we performed classification of biofilm category and biofilm category after the use of different signaling molecules. The algorithm randomly chooses different specimens for training and testing. In each evaluation, it is possible for a random forest classifier to give different results, so we applied cross-validation to assess the overall performance of different data specimens.