# PCA
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
%matplotlib inline
# Veri Seti
url='https://raw.githubusercontent.com/cagriemreakin/Machine-Learning/master/4%20-%20Dimensionality%20Reduction/1%20-%20Principal%20Component%20Analysis/Wine.csv'
dataset = pd.read_csv(url)
X = dataset.iloc[:, 0:13].values
y = dataset.iloc[:, 13].values
# Boyut Sayısı
size=X.shape[1]
# Eğitim ve Test Set Görselleştirme için Kullanılacak
def visualize(X,y,title):
from matplotlib.colors import ListedColormap
plt.figure(figsize=(12,7))
X_set, y_set = X, y
X1, X2 = np.meshgrid(np.arange(start = X_set[:, 0].min() - 1, stop = X_set[:, 0].max() + 1, step = 0.01),
np.arange(start = X_set[:, 1].min() - 1, stop = X_set[:, 1].max() + 1, step = 0.01))
plt.contourf(X1, X2, classifier.predict(np.array([X1.ravel(), X2.ravel()]).T).reshape(X1.shape),
alpha = 0.75, cmap = ListedColormap(('red', 'green', 'blue')))
plt.xlim(X1.min(), X1.max())
plt.ylim(X2.min(), X2.max())
for i, j in enumerate(np.unique(y_set)):
plt.scatter(X_set[y_set == j, 0], X_set[y_set == j, 1],
c = ListedColormap(('red', 'green', 'blue'))(i), label = j)
plt.title('Logistic Regression ' + title)
plt.xlabel('PC1')
plt.ylabel('PC2')
plt.legend()
plt.show()
# Eğitim ve Test Setine Ayırma
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0)
# Feature Scaling
from sklearn.preprocessing import StandardScaler
sc = StandardScaler()
X_train = sc.fit_transform(X_train)
X_test = sc.transform(X_test)
#Component sayısına karar verebilmek için n_components' a parametre olarak None değerini atıyoruz.
#Böylece veri setimizdeki explained variance ve cumulative variance değerlerini bulup grafikte gösteriyoruz.
from sklearn.decomposition import PCA
pca = PCA(n_components = None)
X_train = pca.fit_transform(X_train)
X_test = pca.transform(X_test)
explained_variance = pca.explained_variance_ratio_
cumVar = np.cumsum(np.round(explained_variance, decimals=4)*100)
# Component Sayısına Göre Değişen Explained ve Cumulative Variance Değerini Grafik Üzerinde Gösterme
plt.figure(figsize=(10, 5))
# 100 Üzerinden Explained Variance Değerlerinin Bar ile Gösterimi
plt.bar(range(1,explained_variance.size+1,1), explained_variance*100, align='center', label='Variance', color = 'b')
# 100 Üzerinden Cumulative Variance Değerlerinin Step ile Gösterimi
plt.step(range(1,explained_variance.size+1,1), cumVar, where='mid',label='Cumulative Variance',color='red')
plt.title(' Explained Variance & Cumulative Variance')
plt.ylabel('Explained Variance')
plt.xlabel('Principal components')
plt.legend(loc='best')
plt.show()
Yukarıdaki grafikte 2 component ile toplam variance'ın %60' ı açıklanabilir. 2 boyutlu düzlemde sonuçları göstereceğimiz için 2 seçtik. Fakat araştırmalar, toplam variance'ın en az %70-80 arasında olması gerektiğini söylüyor.
def pca (X_train,X_test,componentNumber):
pca2 = PCA(n_components=componentNumber)
X_train_pca = pca2.fit_transform(X_train)
X_test_pca = pca2.transform(X_test)
return (X_train_pca,X_test_pca)
# Logistic Regression Modeli ile Veri Setini Eğitme
from sklearn.linear_model import LogisticRegression
classifier = LogisticRegression(random_state = 0)
X_train_pca,X_test_pca = pca (X_train,X_test,2)
classifier.fit(X_train_pca, y_train)
LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True, intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1, penalty='l2', random_state=0, solver='liblinear', tol=0.0001, verbose=0, warm_start=False)
classifier2 = LogisticRegression(random_state = 0)
X_train_pca2,X_test_pca2 = pca (X_train,X_test,3)
classifier2.fit(X_train_pca2, y_train)
LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True, intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1, penalty='l2', random_state=0, solver='liblinear', tol=0.0001, verbose=0, warm_start=False)
# Sonuçların Tahmini
y_pred = classifier.predict(X_test_pca)
y_pred2 = classifier2.predict(X_test_pca2)
# Confusion Matrix
from sklearn.metrics import confusion_matrix
cm1 = confusion_matrix(y_test, y_pred)
cm2 = confusion_matrix(y_test, y_pred2)
print("\n Confusion Matrix for 2 Component\n",cm1,"\n\n Confusion Matrix for 3 Component\n",cm2)
Confusion Matrix for 2 Component [[14 0 0] [ 1 15 0] [ 0 0 6]] Confusion Matrix for 3 Component [[14 0 0] [ 0 16 0] [ 0 0 6]]
from sklearn.metrics import accuracy_score
print("n_components = 2 -> Prediciton: %",accuracy_score(y_pred,y_test)*100)
print("n_components = 3 -> Prediction: %",accuracy_score(y_pred2,y_test)*100)
n_components = 2 -> Prediciton: % 97.2222222222 n_components = 3 -> Prediction: % 100.0
visualize(X_train_pca,y_train,'Training Set for 2 Components')
visualize(X_test_pca,y_test,'Test Set for 2 Components')