根据数据集,我们需要使用多类分类算法,利用训练和测试数据对该数据集进行分析。如果我错了请纠正我?

对的。

请告诉我是否使用了此数据集的正确算法。

对。但是应用它们的一个更系统的方法是:首先使用PCA来直观地探索类的可分离性及其组件的相对信息性(您使用的是前两个)。然后,对原高维特征空间和PCA低维特征空间进行了逻辑回归。

#importing the libraries

import numpy as np

import matplotlib.pyplot as plt

import pandas as pd

import seaborn as sns

#importing the Dataset

dataset = pd.read_csv('winequality-red.csv', sep=';') # https://archive.ics.uci.edu/ml/machine-learning-databases/wine-quality/winequality-red.csv

sns.countplot(dataset['quality'])

观察:6类和高级不平衡(6可能是因为我们在您共享的页面中使用了不同的数据集)。

另外,正如我所看到的,我们有9个类,其中这个数据集将被划分。请让我知道如何在不同的类中相应地可视化和绘制数据。

# Feature Scaling

from sklearn.preprocessing import StandardScaler

sc = StandardScaler()

X = sc.fit_transform(X)

#Applying the PCA

from sklearn.decomposition import PCA

fig = plt.figure(figsize=(12,6))

pca = PCA()

pca_all = pca.fit_transform(X)

pca1 = pca_all[:, 0]

pca2 = pca_all[:, 1]

fig.add_subplot(1,2,1)

plt.bar(np.arange(pca.n_components_), 100*pca.explained_variance_ratio_)

plt.title('Relative information content of PCA components')

plt.xlabel("PCA component number")

plt.ylabel("PCA component variance % ")

fig.add_subplot(1,2,2)

plt.scatter(pca1, pca2, c=y, marker='x', cmap='jet')

plt.title('Class distributions')

plt.xlabel("PCA Component 1")

plt.ylabel("PCA Component 2")

量化多类分类性能有许多指标。使用

accuracy

:

# Splitting the dataset into the Training set and Test set

from sklearn.model_selection import train_test_split

#Fiiting the Logistic Regression model to the training set

from sklearn.linear_model import LogisticRegression

from sklearn.metrics import accuracy_score

classifier = LogisticRegression(random_state = 0)

# PCA 2D space

X_train, X_test, y_train, y_test = train_test_split(pd.DataFrame(data=pca_all).iloc[:,0:2], y, test_size = 0.25, random_state = 0)

classifier.fit(X_train, y_train)

y_pred = classifier.predict(X_test)

accuracy_pca_2d = accuracy_score(y_test, y_pred)

# PCA 3D space

X_train, X_test, y_train, y_test = train_test_split(pd.DataFrame(data=pca_all).iloc[:,0:3], y, test_size = 0.25, random_state = 0)

classifier.fit(X_train, y_train)

y_pred = classifier.predict(X_test)

accuracy_pca_3d = accuracy_score(y_test, y_pred)

# PCA 2D space

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.25, random_state = 0)

classifier.fit(X_train, y_train)

y_pred = classifier.predict(X_test)

accuracy_original = accuracy_score(y_test, y_pred)

plt.figure()

sns.barplot(x=['pca 2D space', 'pca 3D space', 'original space'], y=[accuracy_pca_2d, accuracy_pca_3d, accuracy_original])

plt.ylabel('accuracy')

这表明在减少的PCA2D空间中进行分类具有负面影响,至少根据这种测量和设置。

为了可视化混乱矩阵,

this

可以使用。申请原空间案件: