model.gamma = gamma
model.fit(list(data), list(unlab))
score = model.score(list(testData), list(testLabels))
print(score, ' gamma = ', gamma, ' max_iter = ', max_iter, ' alpha = ', alpha)
if (score in scoreDict):
scoreDict[score].append(
'gamma = ' + str(gamma) + ' max_iter = ' + str(max_iter) + ' alpha = ' + str(alpha))
else:
scoreDict[score] = [
'gamma = ' + str(gamma) + ' max_iter = ' + str(max_iter) + ' alpha = ' + str(alpha)]
knnModel = LabelSpreading(kernel='knn')
knnParams = {'n_neighbors': [1, 4, 9, 16], 'max_iter': [10, 100, 1000], 'alpha': [0.2, 0.4, 0.6, 0.8]}
for max_iter in knnParams['max_iter']:
model.max_iter = max_iter
for alpha in knnParams['alpha']:
model.alpha = alpha
for n_neighbors in knnParams['n_neighbors']:
model.n_neighbors = n_neighbors
model.fit(list(data), list(unlab))
score = model.score(list(testData), list(testLabels))
print(score, ' n_neighbors = ', n_neighbors, ' max_iter = ', max_iter, ' alpha = ', alpha)
if (score in scoreDict):
scoreDict[score].append(
'n_neighbors = ' + str(n_neighbors) + ' max_iter = ' + str(max_iter) + ' alpha = ' + str(alpha))
else:
scoreDict[score] = [
'n_neighbors = ' + str(n_neighbors) + ' max_iter = ' + str(max_iter) + ' alpha = ' + str(alpha)]
best = sorted(list(scoreDict.keys()))[-1]
print('Best accuracy: ', best, scoreDict[best])
