def __init__(this): Verifier.__init__(this) # below we add two vectors for comparing between what feature vector is of the ideal Booter # and what feature vector is of the ideal non-Booter. # currently they are defined as completely 0.0 or 1.0; experimenting wtih different values e.g. # averages of Booter training dataset, numbers more often associated with Booters and so on did # not generate consistently better results (sometimes it did, sometimes it didn't). We thus keep # the [1.0,...,1.0] vector as this directly corresponds with the individual distance each element # has to its maximum value. this.vector_booter = [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0]
def __init__(this):
Verifier.__init__(this)
this.p_booter = 0.1001
this.p_non_booter = 0.8999
# calculated: see 'naive_bayes_probabilities.txt'
this.p_booter_characteristics = [
0.97, 0.93, 0.94, 0.37, 0.89, 0.89, 0.38, 0.72, 0.85, 0.74, 0.92, 0.52, 0.22, 0.44, 0.82
]
this.p_non_booter_characteristics = [
0.23, 0.80, 0.67, 0.06, 0.14, 0.62, 0.28, 0.18, 0.35, 0.16, 0.20, 0.19, 0.19, 0.36, 0.66
]
def __init__(this):
Verifier.__init__(this)
# below we add two vectors for comparing between what feature vector is of the ideal Booter
# and what feature vector is of the ideal non-Booter.
# currently they are defined as completely 0.0 or 1.0; experimenting wtih different values e.g.
# averages of Booter training dataset, numbers more often associated with Booters and so on did
# not generate consistently better results (sometimes it did, sometimes it didn't). We thus keep
# the [1.0,...,1.0] vector as this directly corresponds with the individual distance each element
# has to its maximum value.
this.vector_booter = [
1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0,
1.0, 1.0
]
def __init__(this, use_weights = True): Verifier.__init__(this) this.LoadData(use_weights)
