Multiple-Instance Learning (MIL) is a subset of Supervised Learning that is applied when data is available in groups of instances, called bags. A bag receives a positive label if one or more of its instances is positive and a negative label if all of its instances are negative. Due to the representation used in MIL, we can not use traditional Single-Instance Learning (SIL) algorithms and must use a new set of algorithms. MIL is the research area that focuses on the machine learning algorithms that apply to data presented in a Multiple-Instance representation.