This paper describes research on identifying the most significant risk factors for suicide and creating a model for predicting suicide rates. Data sets were collected over which the process of data analysis, processing and merging was performed. The obtained different versions of the data sets were used to train several different versions of the model. Due to the large dimensionality of the data, the reduction of data to the accuracy of predictive models was examined. Risk factors at the global and national levels were found, and the influence of geographical regions, religions and income on risk factors was examined. Predictive models were evaluated, and the results of the most significant risk factors were compared with the results of medical research.