def main():
### PHASE 1 ###
num_classes = 21
name_encoder = preprocessing.LabelBinarizer()
shuffler.shuffleData() # shuffle dataset
trainData = dataset.SigData("data/train_data.csv", name_encoder)
testData = dataset.SigData("data/test_data.csv", name_encoder)
print("created datasets.")
mytrainer = trainer.Trainer(
trainData, testData,
num_users=num_classes) # number of people in the training set
print("training.")
mytrainer.train(num_epochs=60)
exit()
# Load checkpoint as trained weights for CNN
### PHASE 2 ###
# new_dataset for a user
# features = mytrainer.model.get_feature_vectors(new_dataset)
forg_classifier = svm.svc(class_weight='balanced')
forg_classifier_rbf = svm.svc(kernel='rbf', class_weight='balanced')
def get_model(choice='lr', class_weight=None):
if choice == 'svc':
model = svc(verbose=1, class_weight=class_weight, n_jobs=-1)
elif choice == 'lsvc':
model = lsvc(class_weight=class_weight, n_jobs=-1)
elif choice == 'knn':
model = KNeighborsClassifier()
elif choice == 'msvm':
model = MulticlassSVM(C=0.1,
tol=0.01,
max_iter=100,
random_state=0,
verbose=1)
elif choice == 'gnb':
model = gnb(class_weight=class_weight)
elif choice == 'gpc':
model = gpc(class_weight=class_weight)
elif choice == 'sgdc':
model = sgdc(class_weight=class_weight)
elif choice == 'rf':
model = rf(class_weight=class_weight)
# elif choice == 'vw':
# model = vw()
else:
model = lr(class_weight=class_weight)
return model
def quick_ml(df, results, filter_slice, target_df, loss_df, target, loss):
# Create x and y
ind = df['indexes'].loc[filter_slice][0].values
x = results.loc[ind].copy()
y = (target_df.loc[ind, target] < loss_df.loc[ind, loss]).astype(int)
# create sets
x.reset_index(inplace=True, drop=True)
y.index = x.index
row = int(x.shape[0] * .8)
x_train = x.loc[:row]
x_test = x.loc[row:]
y_train = y.loc[:row]
y_test = y.loc[row:]
# scale values
scaler = StandardScaler()
x_train = scaler.fit_transform(x_train)
x_test = scaler.transform(x_test)
# weuights ?
weights = sklearn.utils.class_weight.compute_class_weight(
'balanced', np.array([0, 1]), y_train)
# model
logreg = LogisticRegression(class_weight={0: weights[0], 1: weights[1]})
logreg = MLPClassifier(solver='lbfgs',
alpha=1e-5,
hidden_layer_sizes=(500, 4),
random_state=1)
logreg.fit(x_train, y_train)
predictions = logreg.predict(x_test)
print(classification_report(y_test, predictions))
# Print
logreg = svc()
logreg = KNeighborsClassifier(n_neighbors=2)
logreg = DecisionTreeClassifier(random_state=0)
return None
def train(self, model_id):
#get training status of model container
train_status = model_cont['train_status']
if train_status != "trained":
#load the data to train
data_loader = DataLoader()
dataset = data_loader.load_user_data(user_data_path)
#load model specific parameters
#TODO
if "train_test_split" in params.keys() and params["train_test_split"]:
data_split = params['train_test_split']
trainset = DataProcessor().get_trainset(features, labels, data_split)
#train the model
clf = svc()
clf.fit(dataset['features'], dataset['labels'])
pkl_file = pdumps(clf)
#update the model object with the results of training
model_cont['learned_model']=Binary(pkl_file)
model_cont['train_status'] = "trained"
return model_cont
else:
print("Already trained")
return False
def classifier(data, y, model="forest"):
if model == "forest":
from sklearn.ensemble import RandomForestClassifier as rfc
est = rfc(n_estimators=10, n_jobs=-1)
elif model == "tree":
from sklearn.tree import DecisionTreeClassifier as dtc
est = dtc()
elif model == "extra":
from sklearn.ensemble import ExtraTreesClassifier as etc
est = etc(n_estimators=10, n_jobs=-1)
elif model == "logistic":
from sklearn.linear_model import LogisticRegression as lr
cases = y.nunique()
if cases > 2: est = lr(solver="newton-cg", multi_class="multinomial")
else: est = lr(n_jobs=-1)
elif model == "svm":
from sklearn.svm import SVC as svc
est = svc()
elif model == "boost":
from sklearn.ensemble import GradientBoostingClassifier as gbc
est = gbc(n_estimators=10)
elif model == "neural":
from sklearn.neural_network import MLPClassifier as nnc
est = nnc(max_iter=10, learning_rate_init=1)
est.fit(data, y)
return est
def regression(data, y, model="forest"):
if model == "forest":
from sklearn.ensemble import RandomForestRegressor as rfc
est = rfc(n_estimators=10, n_jobs=-1)
elif model == "tree":
from sklearn.tree import DecisionTreeRegressor as dtc
est = dtc()
elif model == "extra":
from sklearn.ensemble import ExtraTreesRegressor as etc
est = etc(n_estimators=10, n_jobs=-1)
elif model == "linear":
from sklearn.linear_model import LinearRegression as lr
cases = y.nunique()
est = lr(n_jobs=-1)
elif model == "svm":
from sklearn.svm import SVR as svc
est = svc()
elif model == "boost":
from sklearn.ensemble import GradientBoostingRegressor as gbc
est = gbc(n_estimators=10)
elif model == "neural":
from sklearn.neural_network import MLPRegressor as nnc
est = nnc(max_iter=10, learning_rate_init=1)
est.fit(data, y)
return est
def classification(self, metric, folds, alphas, graph):
size = 1.3 * self.report_width // 10
models = {}
models["K nearest neighbors classifier K2"] = knnc(n_neighbors=2)
models["K nearest neighbors classifier K5"] = knnc(n_neighbors=5)
models["K nearest neighbors classifier K10"] = knnc(n_neighbors=10)
models["Decision tree classifier"] = dtc()
models["Logistic classifier"] = logitc()
models["SVM classifier with RBF kernel"] = svc(gamma='scale')
models["SVM classifier with linear kernel"] = svc(kernel='linear')
models["Gaussian naive bayes"] = gnbc()
models["Bernoulli naive bayes"] = bnbc()
models["SGD classifier"] = sgdc(max_iter=10000)
models["Random forest classifier"] = rfc(n_estimators=100)
models["Gradient boosting classifier"] = gbc()
self.models = models
print('\n')
print(self.report_width * '*', '\n*')
print('* CLASSIFICATION RESULTS - BEFORE PARAMETERS BOOSTING \n*')
kf = StratifiedKFold(n_splits=folds, shuffle=True)
results = []
names = []
for model_name in models:
cv_scores = cross_val_score(models[model_name], self.Xt_train, self.yt_train.values.ravel(), cv=kf, error_score=np.nan)
results.append(cv_scores)
names.append(model_name)
print(self.report_width * '*', '')
report = pd.DataFrame({'Classifier': names, 'Score': results})
report['Score (avg)'] = report.Score.apply(lambda x: x.mean())
report['Score (std)'] = report.Score.apply(lambda x: x.std())
report['Score (VC)'] = 100 * report['Score (std)'] / report['Score (avg)']
report.sort_values(by='Score (avg)', inplace=True, ascending=False)
report.drop('Score', axis=1, inplace=True)
display(report)
print('\n')
if graph:
fig, ax = plt.subplots(figsize=(size, 0.5 * size))
plt.title('Classifier Comparison')
#ax = fig.add_subplot(111)
plt.boxplot(results)
ax.set_xticklabels(names)
plt.xticks(rotation=45)
plt.subplots_adjust(hspace=0.0)
plt.show()
return None
def makeSVM(self,**kwargs):
scale_mag,scale_x,scale_y = self.FitScale(self.data)
self.scaled_data = np.c_[scale_mag,scale_x,scale_y]
self.completeness = svc(probability=True,**kwargs)
if self.spatial:
self.completeness.fit(self.scaled_data,self.det.ravel())
else:
self.completeness.fit(np.c_[scale_mag],self.det.ravel())
def makeSVM(self, **kwargs):
scale_mag, scale_x, scale_y = self.FitScale(self.data)
self.scaled_data = np.c_[scale_mag, scale_x, scale_y]
self.completeness = svc(probability=True, **kwargs)
if self.spatial:
self.completeness.fit(self.scaled_data, self.det.ravel())
else:
self.completeness.fit(np.c_[scale_mag], self.det.ravel())
def support_vector_machine():
#Import Library
from sklearn import svm
#Assumed you have, X (predictor) and Y (target) for training data set and x_test(predictor) of test_dataset
# Create SVM classification object
model = svm.svc() # there is various option associated with it, this is simple for classification. You can refer link, for mo# re detail.
# Train the model using the training sets and check score
model.fit(X, y)
# R^2 score
model.score(X, y)
#Predict Output
predicted= model.predict(x_test)
def SVC(test_data, test_label, train_data, train_label, d):
svm_model_poly_classifier = svc(kernel='poly',
degree=d,
C=5,
gamma=0.05,
probability=True)
#It must be one of ‘linear’, ‘poly’, ‘rbf’, ‘sigmoid’, ‘precomputed’ or a callable.
svm_train_error = svm_model_poly_classifier.fit(
train_data, train_label).score(train_data, train_label)
svm_test_error = svm_model_poly_classifier.score(test_data, test_label)
y_predict = svm_model_poly_classifier.predict(test_data)
return y_predict, (1 - svm_train_error) * len(train_data), (
1 - svm_test_error) * len(test_data)
def CrossValidate(self,c_range=np.logspace(-3.0,3.0,7),gamma_range=np.logspace(-3.0,3.0,7)):
param_grid = dict(C=c_range,gamma=gamma_range)
cv = StratifiedShuffleSplit(self.det,n_iter=5,test_size=0.2)
grid = GridSearchCV(svc(kernel='rbf',cache_size=1000),param_grid=param_grid,cv=cv)
scale_mag,scale_x,scale_y = self.Scale(self.data)
if self.spatial:
grid.fit(np.c_[scale_mag,scale_x,scale_y],self.det.ravel())
else:
grid.fit(np.c_[scale_mag],self.det.ravel())
print("The best parameters are %s with a score of %0.2f"
% (grid.best_params_, grid.best_score_))
def train_model(train_X, train_Y, valid_X, valid_Y, hyper_param1):
# Choose a classifier (here, linear SVM)
clf = svc(C=1.0, kernel='rbf', max_iter=1000)
# train
clf.fit(train_X, train_Y)
# validation
valid_Y_hat = clf.predict(valid_X)
accuracy = np.sum((valid_Y_hat == valid_Y)) / 200.0 * 100.0
print 'validation accuracy = ' + str(accuracy) + ' %'
return clf, accuracy
def train(self, model_id):
try:
conn = MongoClient()
print("\nConnection Successful")
mlaas_db = conn[DB_NAME] #connect to db
mlaas_db.authenticate(USERNAME, PASS)
models = mlaas_db[COLL_NAME] #load the collection
#get the model corresponding to the id provided
model_cont = models.find_one({'_id': ObjectId(model_id)})
assert model_cont, "Invalid model ID"
#get training status of model container
train_status = model_cont['train_status']
#block if model is being trained currently
if train_status != "training":
#init the data loader and data processor
data_loader = DataLoader()
data_processor = DataProcessor()
#params = model['parameters']
dataset = data_loader.load_user_data(data_path,
USER_DATA_FNAME)
if 'train_test_split' in params.keys():
data_split = params['train_test_split']
#TODO: implement this function
trainset = data_processor.get_trainset(
features, labels, data_split)
clf = svc()
clf.fit(dataset['features'], dataset['labels'])
save_pkl(clf.coef_, data_path, WEIGHTS_FNAME)
#model_cont['path_to_weights'] = PATH_TO_WEIGHTS
model_cont['train_status'] = "trained"
models.update({'_id': ObjectId(model_id)},
{'$set': model_cont},
upsert=False)
#TODO: figure out why this form of catch clause does not work
except ConnectionFailure as conn_e:
print("\nCould not connect to server. \
Raised the following exception:\n{}".format(conn_e))
def CrossValidate(self,
c_range=np.logspace(-3.0, 3.0, 7),
gamma_range=np.logspace(-3.0, 3.0, 7)):
param_grid = dict(C=c_range, gamma=gamma_range)
cv = StratifiedShuffleSplit(self.det, n_iter=5, test_size=0.2)
grid = GridSearchCV(svc(kernel='rbf', cache_size=1000),
param_grid=param_grid,
cv=cv)
scale_mag, scale_x, scale_y = self.Scale(self.data)
if self.spatial:
grid.fit(np.c_[scale_mag, scale_x, scale_y], self.det.ravel())
else:
grid.fit(np.c_[scale_mag], self.det.ravel())
print("The best parameters are %s with a score of %0.2f" %
(grid.best_params_, grid.best_score_))
def __fitsvm__(self, **args):
train_objects_files = returnFiles(args['train_object_folder'])
train_objects = returnImages(train_objects_files)
train_objects = preproc(train_objects)
svc_args = dict(
zip(getargspec(svc.__init__)[0][1:],
getargspec(svc.__init__)[3]))
classifier_raw = svc(**hlp.chooseArgs(svc_args, self.args))
number_of_objects = hlp.listLengths(train_objects)
shp = np.shape(train_objects)
train_objects = np.reshape(train_objects,
(shp[0] * shp[1], shp[2] * shp[3]))
train_object_labels_expended = hlp.listExpend(
args['train_object_labels'], number_of_objects)
self.data = classifier_raw.fit(train_objects.tolist(),
train_object_labels_expended)
self.threshold = 0.5
if args['is_save']:
self.__save()
def svc(x_train, y_train, C=1, kernel='rbf', x_test=False):
'''
C-Support Vector Classification. based on libsvm. complexity is more than quadratic.
hard to scale to dataset with more than a couple of 10000 samples.
kernel can be:It must be one of ‘linear’, ‘poly’, ‘rbf’, ‘sigmoid’, ‘precomputed’ or a callable.
If a callable is given it is used to pre-compute the kernel matrix from data matrices;
that matrix should be an array of shape (n_samples, n_samples).
:param x_train:
:param y_train:
:param x_test:
:return:
'''
model = svm.svc(C=C, kernel=kernel)
model.fit(x_train, y_train)
if x_test == False:
print('Score:', model.score(x_train, y_train))
return model
else:
return model, model.predict(x_test)
def predictedSVC():
start = time.time()
clf = Pipeline([('vect', vectorizer),
('clf',
svc(tol=1e-3,
verbose=0,
random_state=42,
C=1.0,
max_iter=-1,
gamma='scale',
probability=True))],
verbose=True)
parameters = {
'vect__ngram_range': [(1, 1), (1, 2), (1, 3), (1, 4), (1, 5)],
'clf__tol': (1e-3, 1e-2, 5e-3, 2e-3, 3e-3, 4e-3),
'clf__gamma': ('auto', 'scale'),
'clf__C': (1.0, .1, .2, .3, .4, 0.5, 0.6, 0.7, 0.8, 0.9)
}
gs_clf = GridSearchCV(clf, parameters, cv=5, iid=False, n_jobs=-1)
gs_clf.fit(docs_train, y_train)
y_predicted = gs_clf.predict(docs_test)
print(gs_clf.best_params_)
print("End.......... total=%.2f s" % (start - time.time()))
# Print the classification report
print(
metrics.classification_report(y_test,
y_predicted,
target_names=dataset.target_names))
cm = metrics.confusion_matrix(y_test, y_predicted)
print(cm)
plt.matshow(cm, cmap=plt.cm.jet)
#plt.show()
joblib.dump(gs_clf, "svc_model.plk")
def score(params):
print "Training with params : "
print params
if 'scale' in params:
if params['scale'] == 1:
preprocessing.scale(X_train), preprocessing.scale(
X_test), preprocessing.scale(y_train), preprocessing.scale(
y_test), preprocessing.scale(ou_test)
del params['scale']
# watchlist = [(dvalid, 'eval'), (dtrain, 'train')]
clf = svm.svc(**params)
a = clf.fit(X_train, y_train)
preds = clf.predict(X_test)
print preds
score = Scorer(preds, y_test, ou_test)
print score
return {'loss': score, 'status': STATUS_OK}
print()
print("Test Set Size")
print(Y_test.shape)
print()
print("Classes:")
print(target_names)
print("----------------------")
## Vectorization object
vectorizer = TfidfVectorizer(
strip_accents=None,
preprocessor=None,
)
## classifier
svm = svc()
## With a Pipeline object we can assemble several steps
## that can be cross-validated together while setting different parameters.
pipeline = Pipeline([
('vect', vectorizer),
('svm', svm),
])
## Setting parameters.
## Dictionary in which:
## Keys are parameters of objects in the pipeline.
## Values are set of values to try for a particular parameter.
parameters = {
'vect__tokenizer': [None, stemming_tokenizer],
# Import Library from sklearn import svm # Assumed you have, X (predictor) and Y (target) for training data set and x_test(predictor) of test_dataset X = [[0, 0], [2, 2]] y = [0.5, 2.5] # Create SVM(Support Vector Machine) classification object model = svm.svc(gamma='scale') # Train the model using the training sets and check score model.fit(X, y) model.score(X, y) # Predict Output predicted = model.predict(x_test)
print(svmP.w) # In[9]: #output prime results test2 = svmD.predict(x_test) res2 = cm(y_test, test2) print(res2) print(svmD.b) print(svmD.w) print(svmD.a) # In[10]: #lib prime libP = svc() libP.set_params(dual=False) libP.set_params(C=soft) #svc.set_params(max_iter=10000) libP.fit(x_train, y_train) test3 = libP.predict(x_test) print(cm(y_test, test3)) print(libP.coef_) # In[50]: #lib dual libD = svc() libD.set_params(dual=True) libD.set_params(C=soft) libD.set_params(max_iter=100000)
print 'f1 macro:', res
print
# color = cm(1. * i / NUM_COLORS) # color will now be an RGBA tuple
# cm = plt.get_cmap('gist_rainbow')
# fig = plt.figure(figsize=(8.0, 5.0))
# ax = fig.add_subplot(111)
# # ax.set_color_cycle([cm(1. * i / NUM_COLORS) for i in range(NUM_COLORS)])
# ax.plot(range(len(scores)), scores, label=str(threshold))
# ax.text(len(scores) - 1, scores[len(scores) - 1], threshold, fontsize='smaller')
# plt.show()
print name
return res
vec_list = [tf(), cv()]
clf_list = [svc(), lr()]
threshold_list = np.arange(0.5, 3, 0.5)
print len(threshold_list)
# results_size = (len(vec_list), len(clf_list),len(threshold_list))
# results = np.zeros(results_size, dtype = np.float)
# a, b, c = range(3), range(3), range(3)
# def my_func(x, y, z):
# return (x + y + z) / 3.0, x * y * z, max(x, y, z)
grids = np.vectorize(run)(*np.ix_(threshold_list, vec_list, clf_list))
# mean_grid, product_grid, max_grid = grids
print len(grids)
try:
print grids.shape
except:
print type(grids)
sex = 'F'
scaler = StandardScaler()
data_partial = data[data['Sex'] == sex].drop('Sex', axis=1)
# corr_matrix_f, corr_matrix_m = data_f.corr(), data_m.corr()
# plot_corr_matrices(corr_matrix_f, corr_matrix_m)
y = data_partial['EmoState']
X = scaler.fit_transform(data_partial.drop('EmoState', axis=1))
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.3,
random_state=71)
models = (('DTC', dtc()), ('SVM', svc(C=10)), ('KNN', knc(n_neighbors=10)),
('SGDC', sgdc()), ('GNBC', gnbc()), ('MLPC',
mlpc(max_iter=1000,
learning_rate='adaptive')))
results = []
names = []
seed = 13
scoring = 'accuracy'
for name, model in models:
kfold = model_selection.KFold(n_splits=10, random_state=seed, shuffle=True)
cv_results = model_selection.cross_val_score(model,
X_train,
y_train,
cv=kfold,
scoring=scoring)
y = x['PFLAG'].values
x = x.drop(['PFLAG'],axis =1).values
Y = X['PFLAG'].values
X = X.drop(['PFLAG'], axis =1).values
pca = PCA(n_components = 50)
x= pca.fit_transform(x)
X= pca.transform(X)
Print(x.shape)
#modeling
clf = RandomForestClassifier(n_estimators = 400, max_depth = 4)
clf = LogisticRegression(random_state = 0, solver = 'lbfgs')
clf = svm.svc(kernel = 'linear') #rbf, linear, ploy, sigmoid
clf = AdaBoostClassfier(DecisionTreeClassifier(max_depth =1). algorithm = 'SAMME',n_estimators = 200)
clf. fit(x,y)
pred = np.matrix(clf.predict(X))
correct = pred ==Y
Accuracy = np.sum(correct)/correct.shape[1]
print('Accuracy:' +str(accuracy*100))
#logistic ROC curve and Confustion Matrix
Y = Y.tolist()
pred = pred.T.tolist()
fpr, tpr, thresholds = roc_curve(Y, pred)
plt.plot(fpr)
print.show()
print 'test samples:', len(y_test)
vec = cPickle.load(open('AE_unlabeled_vec2014-11-17 11:56:27.965705'))
# vec = tf(vocabulary=old_vec.vocabulary_)
X_train = vec.transform(train_data)
X_test = vec.transform(test_data)
print X_train.shape
print X_test.shape
# load params
W = cPickle.load(open('W_corr0.3_batchsize20_epochs100'))
b = cPickle.load(open('b_corr0.3_batchsize20_epochs100'))
print 'W:', W.shape, 'b:', b.shape
X_train = get_rep(X_train)
X_test = get_rep(X_test)
clf = svc()
clf.fit(X_train, y_train)
y_pred = clf.predict(X_test)
print_top_features(vec, clf)
print classification_report(y_train, clf.predict(X_train))
print confusion_matrix(y_test, y_pred)
print classification_report(y_test, y_pred)
print f1_score(y_test, y_pred, pos_label=None, average='macro')
scores = cross_val_score(clf, X_train, y_train, cv=5, scoring='f1')
print scores
import numpy as np
import pandas as pd
from sklearn import svm
file='inputtrain.xlsx'
x1=pd.ExcelFile(file)
df1=x1.parse('Sheet1')
df1.apply(pd.to_numeric, errors='ignore')
df1max, df1min = df1.max(), df1.min()
df=(df1-df1min)/(df1max-df1min)
file='targettrain.xlsx'
x2=pd.ExcelFile(file)
df2=x2.parse('Sheet1')
df2.apply(pd.to_numeric, errors='ignore')
file='inputtest.xlsx'
x3=pd.ExcelFile(file)
df3=x3.parse('Sheet1')
df3.apply(pd.to_numeric, errors='ignore')
df3max, df3min = df3.max(), df3.min()
df_test=(df3-df3min)/(df3max-df3min)
df=np.array(df)
df2=np.array(df2)
df_test=np.array(df_test)
#Assumed you have, X (predictor) and Y (target) for training data set and x_test(predictor) of test_dataset
# Create SVM classification object
model = svm.svc(kernel='linear', c=1, gamma=1)
# there is various option associated with it, like changing kernel, gamma and C value. Will discuss more # about it in next section.Train the model using the training sets and check score
model.fit(df, df2)
model.score(df, df2)
#Predict Output
predicted= model.predict(df_test)
from sklearn import datasets from sklearn import svm clf = svm.svc(gamma=0.001, c=100.) digits = datasets.load_digits() clf.fit(digits.data[:-1], digits.target[:-1]) result = clf.predict(digits.data[-1]) pprint(resuli)
x = results.loc[ind].copy()
y = (down.loc[ind, 'd6'] < up.loc[ind, 'u1']).astype(int)
x.reset_index(inplace=True, drop=True)
y.index = x.index
row = int(x.shape[0] * .8)
x_train = x.loc[:row]
x_test = x.loc[row:]
y_train = y.loc[:row]
y_test = y.loc[row:]
scaler = StandardScaler()
x_train = scaler.fit_transform(x_train)
x_test = scaler.transform(x_test)
weights = sklearn.utils.class_weight.compute_class_weight(
'balanced', np.array([0, 1]), y_train)
logreg = LogisticRegression(class_weight={0: weights[0], 1: weights[1]})
logreg = MLPClassifier(solver='lbfgs',
alpha=1e-5,
hidden_layer_sizes=(5, 2),
random_state=1)
logreg.fit(x_train, y_train)
predictions = logreg.predict(x_test)
print(classification_report(y_test, predictions))
logreg = svc()
logreg = KNeighborsClassifier(n_neighbors=2)
logreg = DecisionTreeClassifier(random_state=0)
def classification(self, metric, folds, printt=True, graph=False):
size = self.graph_width
if len(self.y.iloc[:,0].unique()) > 2:
struct = 'multiclass'
else:
struct = 'binary'
# significant model setup differences should be list as different models
models = {}
models["Linear discriminant analysis"] = ldac()
models["Nearest centroid classifier euclidian"] = ncc(metric='euclidean')
models["Nearest centroid classifier manhattan"] = ncc(metric='manhattan')
models["K nearest neighbors classifier K2"] = knnc(n_neighbors=2)
models["K nearest neighbors classifier K5"] = knnc(n_neighbors=5)
models["K nearest neighbors classifier K10"] = knnc(n_neighbors=10)
models["Decision tree classifier"] = dtc()
models["Gaussian naive bayes"] = gnbc()
models["Bernoulli naive bayes"] = bnbc(binarize=0.5)
models["Multinomial naive bayes"] = mnbc()
models["SGD classifier"] = sgdc(max_iter=10000)
models["Ridge classifier"] = rc()
if len(self.Xt_train) < 10000:
models["SVM classifier RBF"] = svc(gamma='scale')
models["SVM classifier Linear"] = svc(kernel='linear')
models["SVM classifier Poly"] = svc(kernel='poly')
if self.Xt_train.shape[0] < 10000 or self.Xt_train.shape[1] < 5:
models["Gradient boosting classifier"] = gbc()
models["Random forest classifier"] = rfc(n_estimators=100)
if struct == 'multiclass':
models["Logistic classifier multinomial"] = logitc(multi_class='multinomial', solver='lbfgs')
models["Logistic classifier auto"] = logitc(multi_class='auto')
models["Logistic One vs Rest"] = ovrc(logitc())
models["Logistic One vs One"] = ovoc(logitc())
if struct == 'binary':
models["Logistic classifier"] = logitc(max_iter=2000)
self.models = models
kf = StratifiedKFold(n_splits=folds, shuffle=True)
results = []
names = []
et = []
for model_name in models:
start = time.time()
cv_scores = cross_val_score(models[model_name], self.Xt_train, self.yt_train, cv=kf, scoring=metric, error_score=np.nan)
results.append(cv_scores)
names.append(model_name)
et.append((time.time() - start))
#print(model_name, time.time() - start)
report = pd.DataFrame({'Model': names, 'Score': results, 'Elapsed Time': et})
report['Score (avg)'] = report.Score.apply(lambda x: x.mean())
report['Score (std)'] = report.Score.apply(lambda x: x.std())
report['Score (VC)'] = 100 * report['Score (std)'] / report['Score (avg)']
report.sort_values(by='Score (avg)', inplace=True, ascending=False)
report.drop('Score', axis=1, inplace=True)
report.reset_index(inplace=True, drop=True)
self.report_performance = report
if printt:
print('\n')
print(self.report_width * '*', '\n*')
print('* CLASSIFICATION RESULTS - BEFORE PARAMETERS BOOSTING \n*')
print(self.report_width * '*', '')
print(report)
print('\n')
if graph:
fig, ax = plt.subplots(figsize=(size, 0.5 * size))
plt.title('Classifier Comparison')
#ax = fig.add_subplot(111)
plt.boxplot(results)
ax.set_xticklabels(names)
plt.xticks(rotation=45)
plt.subplots_adjust(hspace=0.0, bottom=0.25)
self.graphs_model.append(fig)
plt.show()
return None
# Train the model using the training sets and check score model.fit(X, y) model.score(X, y) #Predict Output predicted= model.predict(x_test) #Import Library from sklearn import svm #Assumed you have, X (predic tor) and Y (target) for training data set and x_test(predictor) of test_dataset # Create SVM classification object model = svm.svc() # there is various option associated with it, this is simple for classification. You can refer link, for mo# re detail. # Train the model using the training sets and check score model.fit(X, y) model.score(X, y) #Predict Output predicted= model.predict(x_test) #Import Library from sklearn.naive_bayes import GaussianNB
#train x_train= x[split:] y_train= Y[split:] #test x_test= x[:split] y_test= Y[:split] # In[ ]: #svc from sklearn.svm import svc svc().fit(X,Y) # In[ ]: #support vector cla cls= SVC().fix(X_train , y_train) # In[ ]: #classifier accuracy from sklearn.metrics import accuracy_score accuracy_scorey(y_true, y_predicted)
"""Using Support Vector Machine (SVM) model to prdict the competition test target values
"""
import pandas as pd
from sklearn import cross_validation
from sklearn.svm import SVC as svc
from sklearn.metrics import accuracy_score
training_data = pd.read_csv('../datasets/numerai_training_data.csv')
tournament_data = pd.read_csv('../datasets/numerai_tournament_data.csv')
#this returns four arrays which is in the order of features_train, features_test, labels_train, labels_test
features_train, features_test, labels_train, labels_test = cross_validation.train_test_split(
training_data.iloc[:, 0:21],
training_data['target'],
test_size=0.3,
random_state=0)
clf = svc(C=1.0).fit(features_train, labels_train)
#predicting our target value with the 30% remnant of the training_data
predictions = clf.predict(features_test)
print predictions
accuracy = accuracy_score(predictions, labels_test)
print accuracy
#c = 1.0 -> 0.514361849391
#c = 100.0 -> 0.518133997785
from sklearn import svm
import pickle
import pandas
#dataframe = pandas.read_csv("file name")
#array = dataframe.values
# kernel = 'rbf'
model = svm.svc(kernel='linear', c=1, gamma=1)
model.fit(X_train, y_train)
model.score(X_train, y_train)
#predict
predicted = model.predict(x_test)
#Import Library from sklearn import svm #Assumed you have, X (predic tor) and Y (target) for training data set and x_test(predictor) of test_dataset # Create SVM classification object model = svm.svc() # there is various option associated with it, this is simple for classification. You can refer link, for mo# re detail. # Train the model using the training sets and check score model.fit(X, y) model.score(X, y) #Predict Output
def make_model(col_labels = None, year = 2017, model_type = None):
"""make and run model"""
data = pd.read_csv('NCAA2001_2017.csv')
data_2018 = pd.read_csv('NCAA2018.csv')
data_2018['year'] = 2018
data = data.append(data_2018)
# data to pull from the data frame
if col_labels is None:
col_labels = [
'TopEFGPer', # effective field goal percentage
'TopFTR', # free throw rate
'TopTOPer', # turnover percentage
'TopDRTG', # defensive rating
'TopSOS', # strength of schedule
'BotEFGPer',
'BotFTR',
'BotTOPer',
'BotDRTG',
'BotSOS'
]
# don't scale SeedType
if 'SeedType' in col_labels:
col_labels.remove('SeedType')
if len(col_labels) != 0:
data[col_labels] = scale(data[col_labels])
col_labels.insert(0, 'SeedType')
else:
data[col_labels] = scale(data[col_labels])
# change SeedTypes to integers in case need to encode later
data = data.replace(
['OneSixteen', 'TwoFifteen', 'ThreeFourteen',
'FourThirteen', 'FiveTwelve', 'SixEleven',
'SevenTen', 'EightNine'],
[1, 2, 3, 4, 5, 6, 7, 8])
train = data.loc[(data['year'] != year) & \
(data['year'] != 2018)][col_labels]
train_results = data.loc[(data['year'] != year) & \
(data['year'] != 2018)]['Upset'] # not a df
test = data.loc[data['year'] == year][col_labels]
results_columns = ['SeedType', 'TopSeed', 'BotSeed', 'Upset']
test_results = data.loc[data['year'] == year][results_columns]
# have to one-hot the seeding type if that's in there
if 'SeedType' in col_labels:
enc = OneHotEncoder(categorical_features = [0]) # must be first
train = enc.fit_transform(train).toarray()
test = enc.fit_transform(test).toarray()
else:
train = train.as_matrix()
test = test.as_matrix()
# making the model
if model_type == "forest":
model = rf()
elif model_type == "gbc":
model = gbc()
elif model_type == "svc":
model = svc(probability = True)
else:
model = lm.LogisticRegression()
model.fit(train, train_results.as_matrix())
predictions = model.predict_proba(test)
proba = []
for i in range(len(predictions)):
proba.append(predictions[i][1]) # second column is upset percentage
test_results['UpsetProba'] = proba
test_results = test_results.sort('UpsetProba', ascending = 0)
print(test_results)
model = tree.DecisionTreeClassifier(
criterion='gini'
) # for classification, here you can change the algorithm as gini or entropy (information gain) by default it is gini
# model = tree.DecisionTreeRegressor() for regression
# Train the model using the training sets and check score
model.fit(X, y)
model.score(X, y)
#Predict Output
predicted = model.predict(x_test)
#Support Vector Machine
#Import Library
from sklearn import svm
#Assumed you have, X (predictor) and Y (target) for training data set and x_test(predictor) of test_dataset
# Create SVM classification object
model = svm.svc(
) # there is various option associated with it, this is simple for classification. You can refer link, for mo# re detail.
# Train the model using the training sets and check score
model.fit(X, y)
model.score(X, y)
#Predict Output
predicted = model.predict(x_test)
#Naive Bayes
#Import Library
from sklearn.naive_bayes import GaussianNB
#Assumed you have, X (predictor) and Y (target) for training data set and x_test(predictor) of test_dataset
# Create SVM classification object model = GaussianNB() # there is other distribution for multinomial classes like Bernoulli Naive Bayes, Refer link
# Train the model using the training sets and check score
model.fit(X, y)
#Predict Output
predicted = model.predict(x_test)
#!/usr/bin/env python3 #Import Library from sklearn import svm #Assumed you have, X (predictor) and Y (target) for training data set and x_test(predictor) of test_dataset # Create SVM classification object model = svm.svc() # there is various option associated with it, this is simple for classification. You can refer link, for mo# re detail. # Train the model using the training sets and check score model.fit(X, y) model.score(X, y) #Predict Output predicted= model.predict(x_test)
