Machine learning for decision making in medicine and healthcare

In this research we summarize how machine learning algorithms can be used for decision making that can affect health policies. We present modified ANOVA algorithm for identifying marker leukemia genes that allows deeper examination of genes subsets that can increase the risk of developing leukemia. The algorithm uses the ANOVA, the bootstrap and classification to provide an insight whether a particular group of genes affects cancer development. So, medical practitioners can select a group of leukemia genes, test it by the algorithm and further decide whether to examine the group in medical test or switch a gene in the subset. We also present algorithms to outline factors that affect covid geographical distribution and use of vaccines. Some of our key findings are that leukemia genes can be ranked by importance and in rare cases mutations in less important genes can also lead to leukemia development. In terms of covid, we find that the economic development of a country can be related to the willingness of people to vaccinate.