target = dataset.iloc[:, 30].values
data = dataset.iloc[:, 0:30].values
#print(target)
#print(data)
#machine = RandomForestClassifier(criterion = 'gini', max_depth=80, n_estimators = 11)
depth = [10, 20, 30, 50, 60]
for dep in depth:
machine = RandomForestClassifier(criterion='gini',
max_depth=dep,
n_estimators=11)
acc_score, conf_matrix = run_kfold(4, data, target, machine)
print('\n\n')
print('depth = ', str(dep))
print(acc_score)
for i in conf_matrix:
print(i)
'''
depth = 80
accuracy_score:
[0.90474, 0.9046, 0.90586, 0.90568]
confusion_matrix:
[[ 258 429 5 0]
[ 39 33098 1105 3]
[ 1 2342 7749 230]
[ 0 8 601 4132]]
import kfold_template
import pandas as pd
from sklearn import linear_model
dataset = pd.read_csv('regression_dataset.csv')
# y1 is continuous Y, y2 is binomial. here we use y1
#target = dataset.iloc[:,1].values
target = dataset.iloc[:, 2].values
data = dataset.iloc[:, 3:].values
machine = linear_model.LogisticRegression()
acc_score, conf_matr = kfold_template.run_kfold(4, data, target, machine)
print(acc_score, conf_matr)
# print readable results
for item in conf_matr:
print(item)
import kfold_template
from matplotlib import pyplot as plt
from sklearn.datasets.samples_generator import make_blobs
from sklearn import svm
# Data generating
data, target = make_blobs(n_samples=400,
centers=2,
cluster_std=1,
random_state=0)
print(data)
print(target)
plt.scatter(data[:, 0], data[:, 1], c=target)
plt.savefig("plot.png")
r2_scores, accuracy_scores, confusion_matrices = kfold_template.run_kfold(
5, data, target, svm.SVC(kernel="linear"), 1, 1)
print(r2_scores)
print(accuracy_scores)
for i in confusion_matrices:
print(i)
import kfold_template
from sklearn import linear_model
import pandas
dataset = pandas.read_csv("regression_dataset.csv")
target = dataset.iloc[:, 2].values
data = dataset.iloc[:, 3:9].values
machine = linear_model.LogisticRegression()
results_accuracy, results_confusion = kfold_template.run_kfold(
4, data, target, machine)
print(results_accuracy)
for i in results_confusion:
print(i)
import kfold_template
import pandas
# from sklearn import tree
from sklearn.ensemble import RandomForestClassifier
dataset = pandas.read_csv("dataset.csv")
target = dataset.iloc[:,30].values
data = dataset.iloc[:,0:30].values
# print(target)
# print(data)
machine = RandomForestClassifier(n_estimators=21, criterion="gini", max_depth=30)
r2_scores, accuracy_scores, confusion_matrices = kfold_template.run_kfold(3, data, target, machine, 1, 1)
print(r2_scores)
print(accuracy_scores)
for i in confusion_matrices:
print(i)
sub_dataset = dataset.sample(n=1000)
#print(sub_dataset)
target = sub_dataset.iloc[:, 2].values
data = sub_dataset.iloc[:, [5, 6, 7, 8, 12, 16]].values
target = target.astype(int)
max_depth_list = [10]
n_estimators_list = [11]
results_list = []
for i in max_depth_list:
machine = RandomForestClassifier(n_estimators=11,
criterion='gini',
max_depth=i)
accuracy_score, confusion_matrix = kfold_template.run_kfold(
4, data, target, machine)
results_list.append(['Random Forest ', accuracy_score, str(i)])
for i in confusion_matrix:
print(i)
for k in n_estimators_list:
machine = RandomForestClassifier(n_estimators=k,
criterion='gini',
max_depth=10)
accuracy_score, confusion_matrix = kfold_template.run_kfold(
4, data, target, machine)
results_list.append(['Random Forest ', accuracy_score, str(k)])
for i in confusion_matrix:
print(i)
results = pandas.DataFrame(results_list)
#plt.scatter(data[:,0], data[:,1], c=target)
#plt.savefig('plot_circles.png')
'''
try to use higher dimension and linear kernel,
it will give you more accurate result than using non-linear kernel directly.
'''
## [1.0, 0.99, 0.99, 1.0, 1.0] after adding 3rd dimension, we have much higher
## score with linear kernel
machine = svm.SVC(kernel='linear', tol = 0.000001)
accu_score, conf_matrix = kfold_template.run_kfold(5, data, target, machine)
print(accu_score)
for i in conf_matrix:
print(i)
import kfold_template
import pandas
from sklearn import linear_model
dataset = pandas.read_csv("logistic_dataset.csv")
target = dataset.iloc[:, 2].values
data = dataset.iloc[:, 3:9].values
r2_scores, accuracy_scores, confusion_matrices = kfold_template.run_kfold(
5, data, target,
linear_model.LogisticRegression(multi_class="auto", solver="lbfgs"), 1, 1)
print(r2_scores)
print(accuracy_scores)
for confusion_matrix in confusion_matrices:
print(confusion_matrix)
machine = linear_model.LogisticRegression(multi_class="auto", solver="lbfgs")
machine.fit(data, target)
X = [
[24, 55, 31, 3, 0, 7],
[40, 50, 2, 5, 1, 8],
[3, 95, 37, 3, 1, 15],
]
results = machine.predict(X)
print(results)
