【PYTHON】algorithm compare all classification models

 from pandas import read_csv

from matplotlib import pyplot

from sklearn.model_selection import KFold

from sklearn.model_selection import cross_val_score

from sklearn.linear_model import LogisticRegression

from sklearn.tree import DecisionTreeClassifier

from sklearn.neighbors import KNeighborsClassifier

from sklearn.discriminant_analysis import LinearDiscriminantAnalysis

from sklearn.naive_bayes import GaussianNB

from sklearn.svm import SVC

# load dataset

filename = 'pima-indians-diabetes.csv'

names = ['preg', 'plas', 'pres', 'skin', 'test', 'mass', 'pedi', 'age', 'class']

dataframe = read_csv(filename, names=names)

array = dataframe.values

X = array[:,0:8]

Y = array[:,8]

# prepare models

models = []

models.append(('LR', LogisticRegression(solver='liblinear')))

models.append(('LDA', LinearDiscriminantAnalysis()))

models.append(('KNN', KNeighborsClassifier()))

models.append(('CART', DecisionTreeClassifier()))

models.append(('NB', GaussianNB()))

models.append(('SVM', SVC(gamma='auto')))

# evaluate each model in turn

results = []

names = []

scoring = 'accuracy'

for name, model in models:

  kfold = KFold(n_splits=10, random_state=7)

  cv_results = cross_val_score(model, X, Y, cv=kfold, scoring=scoring)

  results.append(cv_results)

  names.append(name)

  msg = "%s: %f (%f)" % (name, cv_results.mean(), cv_results.std())

  print(msg)

# boxplot algorithm comparison

fig = pyplot.figure()

fig.suptitle('Algorithm Comparison')

ax = fig.add_subplot(111)

pyplot.boxplot(results)

ax.set_xticklabels(names)

pyplot.show()