Currently, the amount of drugs available is very large. Its therapeutic importance is indisputable. Although its production is performed on safety and security criteria, the taking of some drugs may involve risks, since the occurrence of Adverse Drug Reactions (ADRs) may have serious consequences for the health of the population.
ADRs are a significant public health concern, because they can potentiate serious injury and even lead to mortality of individuals. The World Health Organization (WHO) defines ADR as "any harmful or undesired effect which manifests itself after drug administration at doses normally used in man for the prophylaxis, diagnosis or treatment of a disease".
The development of predictive procedures would help health professionals, since they could avoid many unwanted and unknown serious effects before marketing the drug. Various computer methods are known, the main objective of which is to predict unknown drug adverse effects. Predictive Data Mining models can be a valuable aid to this.
According to Fayyad et al., "Data Mining (DM) consists in the accomplishment of data analysis and the application of discovery algorithms that, under certain computational limitations, produce a set of patterns of certain data".
​
About
