Features are metric
Features are continous or ordinal
There is r>0.3 correlation between the features in the dataset
We have over 100 observations and at least 5 observations per feature
Sample is homogeneous
(c)ArmyofQuants