result = pool.map(proc, enumerate(train_codes))
ordered = zip(*sorted(result))
H_train = ordered[1]
H_test = ordered[2]
#print H_train
# Compare distance of code to codes and classify
print "Training accuracy:", evaluate(H_train, train_labels, test_codes)
print "Testing accuracy:", evaluate(H_test, test_labels, codes)
elif args.m == 'VOTE':
# Run SMO on each label pair {i=-1, j=+1}
NUM_LABELS = 10
m = len(train_data)
models = {}
G = gram(train_data)
# Simple voting, no tie-resolution
votes_simple = np.zeros((m, NUM_LABELS))
for i in range(NUM_LABELS):
models[i] = {}
for j in range(i+1, NUM_LABELS):
print i,j
model = Model(i, j, G)
models[i][j] = model
model.train(train_data, train_labels)
for k in range(m):
v = model.test(k)
votes_simple[k][v] += 1
ties = 0
h_simple = np.zeros(m)
data_file = "../data/spambase/spambase.data"
dmat = []
f = open(data_file, "r")
for line in f:
x = line.split(',')
x = [float(e) for e in x]
dmat.append(x)
data = np.array(dmat)
# k-folds xvalidation
k = 10
kfolder = KFolder(data, k, standard=True, shuffle=True)
for i in range(k-1):
print "Fold:", i+1
# Get data and labels at fold k
X,Y = kfolder.testing(i+1)
# Get the testing data
Xi,Yi = kfolder.testing(i)
Yi[Yi==0] = -1.0
# Train
Y[Y==0] = -1.0
G, Gi = gram(X), tgram(X, Xi)
a, b = train(X, Y.ravel(), G, C=1e-4, tol=1e-4, eps=1e-3)
# Test
print "Training accuracy:", test(Y, Y, G, a, b)
print "Testing accuracy:", test(Y, Yi, Gi, a, b)