def test():
"""
test Perceptron, SVM and LDA accuracy
"""
tested_models = [
TestedModel('Perceptron', models.Perceptron()),
TestedModel('SVM', models.SVM()),
TestedModel('LDA', models.LDA()),
]
k = 10000
iterations = 500
for i, m in enumerate(MS):
for j in range(iterations):
X, y = sample_d(m)
X_t, y_t = sample_d(k)
for tested in tested_models:
tested.model.fit(X, y)
score = tested.model.score(X_t, y_t)
tested.add_accuracy(m, score['accuracy'])
plt.figure()
for tested in tested_models:
plt.plot(MS, [tested.accuracy[m] for m in MS],
marker='.',
label=tested.name)
plt.legend()
plt.title('Training batch size vs. accuracy')
plt.xlabel('m')
plt.ylabel('accuracy')
plt.show()
for tested in tested_models:
print(tested.name, tested.accuracy)
def main(args):
"""
Fit a model's parameters given the parameters specified in args.
"""
# Load the data.
X, vocab = build_dtm(args.datadir, args.num_documents, args.max_words)
# build the appropriate model
inference = models.MeanFieldVariationalInference(
num_topics=args.num_topics,
num_docs=X.shape[0],
num_words=X.shape[1],
alpha=args.alpha,
beta=args.beta,
epsilon=args.epsilon)
# Run the inference method
model = models.LDA(inference=inference)
model.fit(X=X,
iterations=args.num_vi_iterations,
estep_iterations=args.num_estep_iterations)
# predict topic assignments for words in corpus
preds = model.predict(vocab=vocab, K=args.top_k)
# output model predictions
with open(args.predictions_file, 'w') as file:
for pred in preds:
file.write(pred + '\n')
def main(args):
"""
Fit a model's parameters given the parameters specified in args.
"""
# Load the data.
X, vocab = build_dtm(args.data, args.num_documents)
# build the appropriate model
if args.inference.lower() == 'gibbs-sampling':
inference = models.GibbsSampling(num_topics=args.num_topics,
num_docs=X.shape[0],
num_words=X.shape[1],
alpha=args.alpha,
beta=args.beta)
elif args.inference.lower() == 'sum-product':
inference = models.SumProduct(num_topics=args.num_topics,
num_docs=X.shape[0],
num_words=X.shape[1],
num_nonzero=X.nnz,
alpha=args.alpha,
beta=args.beta)
else:
raise Exception(
'The method given by --inference is not yet supported.')
# Run the inference method
model = models.LDA(inference=inference)
model.fit(X=X, iterations=args.iterations)
# predict topic assignments for words in corpus
preds = model.predict(vocab=vocab, K=args.top_k)
# output model predictions
with open(args.predictions_file, 'w') as file:
for pred in preds:
file.write(pred + '\n')
pca.fit(data)
new_data2 = pca.fit_transform(data)
new_test_data2 = pca.fit_transform(test_data)
print('pca dimension fit')
'''
svd.fit(data)
print('SVD fit')
new_data = svd.fit_transform(data)
new_test_data = svd.fit_transform(test_data)
print('dimensions transformed')
print(new_data.shape)
lda_model = models.LDA()
s_lda = LinearDiscriminantAnalysis(solver = 'eigen')
s_lda.fit(new_data,y)
s_lda.fit(new_data2,y)
lda_model.fit(new_data,y)
print('model fit')
pred = lda_model.predict(new_test_data)
print(lda_model.score(np.array(pred),test_y))
print(s_lda.score(new_test_data,test_y))
clf = LinearSVC(random_state=0, tol=1e-5)
clf.fit(new_data,y)
# Selecting the best models for classification.
# May 2019
import sys
sys.path.insert(0, '/Users/chirathhettiarachchi/tensorflow/clardia/preprocess')
import experiments as exp
import models as model
import json
import warnings
warnings.filterwarnings("ignore", category=DeprecationWarning)
warnings.filterwarnings("ignore", category=FutureWarning)
warnings.filterwarnings("ignore", category=RuntimeWarning)
warnings.filterwarnings("ignore", category=UserWarning)
# Read selected features
with open('../cache/selected_features.json') as f:
data = json.load(f)
features = data['diabetes']
print("Selected Features: ", features)
x, y = exp.experiment_3(features)
cutoff = x.shape[0] - 30
valx = x[cutoff:]
valy = y[cutoff:]
x = x[:cutoff]
y = y[:cutoff]
model = model.LDA(x, y, x, y, 4)