def SelectAtt(examples):
""" select the best attribute
to split the examples,
Based on IG
"""
length = len(examples[0])
p = [IG(examples, i + 1) for i in range(length - 1)]
ans = ut.getMaxPos(p)
#print p[ans[0]]
return ans[0] + 1
def SelectAtt(examples, featureSet = []):
""" select the best attribute
to split the examples,
Based on IG
"""
if len(featureSet) == 0:
featureSet = range(len(examples[0]) - 1)
p = [IG(examples, i + 1) for i in featureSet]
ans = ut.getMaxPos(p)
#print p[ans[0]]
return ans[0] + 1
def getThreshold(egs, att):
"""
given egs and attribute,
choose the `best` threshold for this att to construct binary decision tree
"""
if egs == []:
return 0
tmp = [eg[att] for eg in egs]
mn = min(tmp)
delta = (max(tmp) - min(tmp)) / N
tmp = [ig(egs, att, i * delta + mn) for i in range(N)]
return ut.getMaxPos(tmp)[0] * delta + mn
def cross_validation(crossData, K = 3):
"""
given a cross training set
return the error rate
"""
num_cross = len(crossData)
totError = 0
totNum = 0
for i in range(num_cross):
f = leastSquare(crossData[i][0])
for x in crossData[i][1]:
totNum = totNum + 1
if ut.getMaxPos(f(x[1:]).tolist())[0] != x[0] - 1:
totError = totError + 1
return (totError + 0.) / totNum
def training(self, trainData, mode = 'diag', subSpaceDim = 2):
# num of classes
K = len(trainData)
# num of features
D = len(trainData[1][0])
#
mk = {}
for k in range(1, K + 1):
mk[k] = ut.getMean(trainData[k], 'array')
#
m = ut.getMean(reduce(ut.add, [trainData[k] for k in range(1, K + 1)]))
#
f = lambda k: len(trainData[k]) * np.dot( mk[k] - m, (mk[k] - m).transpose() )
S_B = reduce(ut.add, [f(k) for k in range(1, K - 1)])
# construct S_W
S_W = []
Sk = range(K + 1)
if mode == 'diag':
S_W = np.diag([1] * D)
else:
g = lambda x, y: np.dot((x - y), (x - y).transpose())
for k in range(1, K + 1 ):
Sk[k] = reduce(ut.add, [g(ut.ls2Vec(x), mk[k]) for x in trainData[k]])
S_W = reduce(ut.add, Sk[1:])
#
[egs, vecs] = np.linalg.eigh(np.dot(np.linalg.inv(S_W), S_B))
pos = ut.getMaxPos(egs, subSpaceDim)
W = ut.arrayExtract(vecs, pos)
# now we will construct trainingData set in the
# sub-feature space, then run Gaussian generative
# method to train the new trainData set to get a
# a classifier
def projectFunc(v):
tmp = np.dot(ut.ls2Vec(v).transpose(), W).tolist()
return tmp[0]
for k in range(K):
trainData[k + 1] = map(projectFunc, trainData[k + 1])
self.projectFunc = projectFunc
self.func = GaussianGM.GaussianGM(trainData)
def f(w, w0, x):
tmp = [np.dot(ww.transpose(), ut.ls2Vec(x))[0, 0] + ww0 for (ww, ww0) in zip(w, w0)]
tmp = ut.getMaxPos(tmp)
return tmp[0] + 1