Statistical Tools for Discovery and Limits

The talk will review statistical methods used in Particle Physics
for establishing discovery and setting limits on model parameters,
with emphasis on current practice at the Large Hadron Collider.  
The talk will address progress in designing searches for new 
phenomena using techniques imported from machine learning,
and asymptotic methods for quantifying discovery significance.  
Particular attention is placed on treatment of systematic 
uncertainties through introduction of nuisance parameters.