def k_nearst_neighbour_classify(self):
knn = KNNClassifier()
predictions = knn.predict_classification(self.X_train, self.y_train,
self.X_test)
accuracy = str(self.main.accuracy(self.y_test, predictions))
accuracy = accuracy[0:4]
self.knn_acc_label.setText(accuracy)
def test_knn(self):
from classifiers import KNNClassifier
if type(self.k) == int:
k = "%s" % self.k
else:
k = "-".join([str(i) for i in self.k])
print("KNNClassifier")
print("---" * 45)
print("Train num = %s" % self.train_num)
print("Test num = %s" % self.test_num)
print("K = %s" % k)
knn = KNNClassifier(self.train_data, self.train_labels, k=self.k, best_words=self.best_words)
classify_labels = []
print("KNNClassifiers is testing ...")
for data in self.test_data:
classify_labels.append(knn.classify(data))
print("KNNClassifiers tests over.")
filepath = "f_runout/KNN-%s-train-%d-test-%d-f-%d-k-%s-%s.xls" % \
(self.type,
self.train_num, self.test_num,
self.feature_num, k,
datetime.datetime.now().strftime(
"%Y-%m-%d-%H-%M-%S"))
self.write(filepath, classify_labels)
def match_codes():
"""Matching full dataset in parallel with customizable embedding method"""
# 0.398657046863 hard, 0.275397642306 soft
working_directory = './data/'
data_codes, data_descriptions = get_data_to_match('slim')
official_codes, official_descriptions = get_official_data()
level = 1
model = HashingEmbedder(
level=level, analyzer='char_wb', ngram_range=(4, 5), norm='l2'
) # word2vecEmbedder() # word2vecEmbedder() # HashingEmbedder() # [HashingEmbedder(level=level, analyzer='char', ngram_range=(3,5), norm='l2')] #[HashingEmbedder(level=level, analyzer='char', ngram_range=(2,3))]
model.embed_data(data_descriptions)
print 'loaded and embedded data'
# test_nNN(model, data_descriptions, data_codes)
official_code_labels = None
true_data_codes = None
use_section = False
if use_section:
official_code_labels = get_section_codes(model.official_codes)
true_data_codes = get_section_codes(data_codes)
else:
official_code_labels = coarsen_codes(model.official_codes)
true_data_codes = coarsen_codes(data_codes)
nNN = 4
classifier = KNNClassifier(n_neighbors=nNN)
classifier.fit(model.official_embeddings, official_code_labels)
pred_codes = classifier.predict(
model.data_embeddings, pbar=True
) # classifier.predict_with_edit_dist(model.data_embeddings, data_descriptions, model.official_descriptions)
errors = pred_codes - true_data_codes
print 'Correctly predicted', 1.0 * np.sum(
errors == 0) / errors.shape[0], 'percent of top level codes'
# plot_confusion_matrix(true_data_codes, pred_codes)
model = word2vecEmbedder()
model.embed_data(data_descriptions)
classifier = KNNClassifier(n_neighbors=nNN)
classifier.fit(model.official_embeddings, official_code_labels)
pred_codes = classifier.predict(
model.data_embeddings
) # classifier.predict_with_edit_dist(model.data_embeddings, data_descriptions, model.official_descriptions)
def classify(data, classifier, num_classes, train_labels, train_features,
test_labels, test_features):
"""
Function used by FoldRunner to execute classification based on the current classifier
:param data: the configuration dictionary
:param classifier: current classifier (from the classifiers list in the configuration file)
:param num_classes: the number of distinct labels (binary or multiclass classification)
:param train_labels: the labels of all train instances
:param train_features: the features of all train instances
:param test_labels: the labels of all test instances
:param test_features: the features of all test instances
:return: the confusion matrix of the classification
"""
if classifier == "NN_keras":
return nnk.classify(data, num_classes, train_labels, train_features,
test_labels, test_features)
elif classifier == "NN_scikit-learn":
return nns.classify(train_labels, train_features, test_labels,
test_features)
elif classifier == "KNN":
return knn.classify(data, train_labels, train_features, test_labels,
test_features)
elif classifier == "NaiveBayes":
return nb.classify(train_labels, train_features, test_labels,
test_features)
elif classifier == "RandomForest":
return rf.classify(train_labels, train_features, test_labels,
test_features)
elif classifier == "LogisticRegression":
return lr.classify(train_labels, train_features, test_labels,
test_features)
def classify_custom_input(self, custom_input_vector):
nb = NaiveBayesClassifier()
nb.train(self.X_train, self.y_train)
prediction = nb.predict([custom_input_vector])
self.custom_text_nb_label.setText(str(prediction[0]))
knn = KNNClassifier()
prediction = knn.predict_classification(self.X_train, self.y_train,
[custom_input_vector])
self.custom_text_knn_label.setText(str(prediction[0]))
rf = SklearnRandomForest()
prediction = rf.random_forest(self.X_train, self.y_train,
[custom_input_vector])
self.custom_text_dt_label.setText(str(prediction[0]))
dt = SklearnDecisionTree()
prediction = dt.decision_tree(self.X_train, self.y_train,
[custom_input_vector])
self.custom_text_rf_label.setText(str(prediction[0]))
def test_nNN(model, data_descriptions, data_codes, nNNmin=2, nNNmax=10):
for nNN in xrange(nNNmin, nNNmax + 1):
classifier = KNNClassifier(n_neighbors=nNN)
t1 = time()
classifier.fit(model.official_embeddings,
get_section_codes(model.official_codes))
# classifier.fit(model.official_embeddings, coarsen_codes(model.official_codes))
pred_section_codes = classifier.predict_with_edit_dist(
model.data_embeddings, data_descriptions,
model.official_descriptions)
# true_coarse_codes = coarsen_codes(data_codes) # .reshape((-1,1))
# errors = pred_codes - true_coarse_codes
true_section_codes = get_section_codes(data_codes) # .reshape((-1,1))
errors = pred_section_codes - true_section_codes
print '------------------------------'
print 'nNN:', nNN
print 'Correctly predicted', 1.0 * np.sum(
errors ==
0) / errors.shape[0], 'percent of top level codes w/ edit dist kNN'
print 'Took', time() - t1, 'seconds'
t1 = time()
# pred_codes = classifier.predict(model.data_embeddings)
# errors = pred_codes - true_coarse_codes
pred_section_codes = classifier.predict(model.data_embeddings)
errors = pred_section_codes - true_section_codes
print 'Correctly predicted', 1.0 * np.sum(
errors ==
0) / errors.shape[0], 'percent of top level code w/ euclidean kNN'
print 'Took', time() - t1, 'seconds'
print '------------------------------'
def test_knn(self):
from classifiers import KNNClassifier
if type(self.k) == int:
k = "%s" % self.k
else:
k = "-".join([str(i) for i in self.k])
print("KNNClassifier")
print("---" * 45)
print("Train num = %s" % self.train_num)
print("Test num = %s" % self.test_num)
print("K = %s" % k)
# print self.train_data
print (self.train_labels)
print (len(self.train_data))
print (self.train_data[0])
knn = KNNClassifier(self.train_data, self.train_labels, k=self.k, best_words=self.best_words)
classify_labels = []
print("KNNClassifiers is testing ...")
for data in self.test_data:
classify_labels.append(knn.classify(data))
print("KNNClassifiers tests over.")
filepath = "f_runout/KNN-%s-train-%d-test-%d-f-%d-k-%s-%s.xls" % \
(self.type,
self.train_num, self.test_num,
self.feature_num, k,
datetime.datetime.now().strftime(
"%Y-%m-%d-%H-%M-%S"))
self.write(filepath, classify_labels)
def KNNClassifier(self,
dataset=None,
class_column=None,
name=None,
pipeline=None,
K=5,
kernel="euclidean",
algo="auto",
weights="uniform",
kernel_params={}):
return KNNClassifier(dataset=dataset,
class_column=class_column,
name=name,
pipeline=pipeline,
K=K,
kernel=kernel,
algo=algo,
weights=weights,
kernel_params=kernel_params,
client=self.client)
localAreaDensity=-1, # Using numActiveColumnsPerInhArea
numActiveColumnsPerInhArea=64,
# All input activity can contribute to feature output
stimulusThreshold=0,
synPermInactiveDec=synPermDec,
synPermActiveInc=synPermInc,
synPermConnected=synPermConn,
maxBoost=1.0,
seed=1956, # The seed that Grok uses
spVerbosity=1)
# Instantiate the spatial pooler test bench.
tb = VisionTestBench(sp)
# Instantiate the classifier
clf = KNNClassifier()
# Train the spatial pooler on trainingVectors.
numCycles = tb.train(trainingVectors, trainingTags, clf, maxTrainingCycles,
minAccuracy)
# Save the permanences and connections after training.
#tb.savePermanences('perms.jpg')
#tb.showPermanences()
#tb.showConnections()
# Get testing images and convert them to vectors.
testingImages, testingTags = data.getImagesAndTags(testingDataset)
testingVectors = encoder.imagesToVectors(testingImages)
# Reverse the order of the vectors and tags for testing
if len(sys.argv) < 2:
terminate()
else:
mode = sys.argv[1]
if mode not in func_mode_list:
terminate()
def show_plot_sample():
fig = plt.figure(figsize=(8, 8))
fig.subplots_adjust(left=0, right=1, bottom=0, top=1, hspace=0.05, wspace=0.05)
for i in tqdm(range(25)):
id = random.randint(0, len(testX) - 1)
images = np.reshape(testX[id], [28, 28])
ax = fig.add_subplot(5, 5, i + 1, xticks=[], yticks=[])
ax.imshow(images, cmap=plt.cm.binary, interpolation='nearest')
ax.text(0, 2, "label:" + str(testY[id]))
ax.text(0, 4, "predict:" + str(knn.predict(testX[id])))
plt.show()
if __name__ == '__main__':
trainX, trainY, testX, testY = load_mnist()
knn = KNNClassifier(train_data=trainX, train_labels=trainY, ord=2)
if mode == 'run_sample':
show_plot_sample()
else:
knn.test_acc(test_data=testX, test_label=testY, K=1)
