
Five models by the Basel Committee for computation of operational risk 159
䊏 Uncover affinities among data consisting of one or more variables, and
䊏 Understand the extent to which the presence of specified variables imply the
presence of other variables across a data pattern.
Both points are important in tracking and analysing risk events. Let’s remember,
however, that datamining is not an end, but a means to an end. Its benefits accrue
from the operational results and the assistance these provide in achieving a specific
objective. The contribution of datamining is that of a discovery-driven approach,
almost always with no a priori hypothesis stated ...