Results of Data Analysis

This chapter includes results of data analysis. The relationship between personality profiles and drug consumption is described and the individual drug consumption risks for different drugs is evaluated. Significant differences between groups of drug users and non-users are identified. Machine learning algorithms solve the user/non-user classification problem for many drugs with impressive sensitivity and specificity. Analysis of correlations between use of different drugs reveals existence of clusters of substances with highly correlated use, which we term correlation pleiades. It is proven that the mean profiles of users of different drugs are significantly different (for benzodiazepines, ecstasy, and heroin). Visualisation of risk by risk maps is presented. The difference between users of different drugs is analysed and three distinct types of users are identified for benzodiazepines, ecstasy, and heroin.