数据降维——LLE模型

模型原型
class sklearn.manifold.LocallyLinearEmbedding(n_neighbors=5,n_components=2,reg=0.001,eigen_solver=’auto’,tol=0,max_iter=300, method=’standard’,hessian_tol=0.0001,modified_tol=1e.12, neighbors_algorithm=’auto’,random_state=None)
参数

• n_neighbors
• n_components
• reg:正则化项的系数
• eigen_solver
• tol
• max_iter
• method:用于指定LLE算法的形式
  
  • ’standard’:使用标准的LLE算法
  • ‘hessian’:使用Hessian eignmap算法
  • ‘modified’:使用modified LLE算法
  • ‘ltsa’:使用local tangent space alignment算法
• hessian_tol:用于method=’hessian’时收敛的阙值
• modified_tol:用于method=’modified’时收敛的阙值
• neighbors_algorithm
• random_state

属性

• embeddingvectors:给出了原始数据在低维空间的嵌入矩阵
• reconstructionerror:给出了重构误差

方法

• fit(X[,y,init])
• transform(X)
• fit_transform(X,[,y])

import numpy as np
import matplotlib.pyplot as plt
from sklearn import datasets,decomposition,manifold

加载数据

def load_data():
    iris=datasets.load_iris()
    return iris.data,iris.target

使用LocallyLinearEmbedding类

def test_LocallyLinearEmbedding(*data):
    X,y=data
    for n in [4,3,2,1]:
        lle=manifold.LocallyLinearEmbedding(n_components=n)
        lle.fit(X,y)
        print('reconstruction_error(n_components=%d):%s'%(n,lle.reconstruction_error_))

X,y=load_data()
test_LocallyLinearEmbedding(X,y)

降维后样本的分布图

def plot_LocallyLinearEmbedding(*data):
    X,y=data
    Ks=[1,5,25,y.size-1]
    fig=plt.figure()
    for i,k in enumerate(Ks):
        lle=manifold.LocallyLinearEmbedding(n_components=2,n_neighbors=k)
        X_r=lle.fit_transform(X)
        ax=fig.add_subplot(2,2,i+1)
        colors=((1,0,0),(0,1,0),(0,0,1),(0.5,0.5,0),(0,0.5,0.5),(0.5,0,0.5),
               (0.4,0.6,0),(0.6,0.4,0),(0,0.6,0.4),(0.5,0.3,0.2),)
        for label,color in zip(np.unique(y),colors):
            position=y==label
            ax.scatter(X_r[position,0],X_r[position,1],label='target=%d'%label,color=color)
        ax.set_xlabel('X[0]')
        ax.set_ylabel('X[1]')
        ax.legend(loc='best')
        ax.set_title("k=%d"%k)
    plt.suptitle('LocallyLinearEmbadding')
    plt.show()

plot_LocallyLinearEmbedding(X,y)

将原始数据的特征直接压缩到一维

def plot_LocallyLinearEmbedding_k_d1(*data):
    X,y=data
    Ks=[1,5,25,y.size-1]
    fig=plt.figure()
    for i,k in enumerate(Ks):
        lle=manifold.LocallyLinearEmbedding(n_components=2,n_neighbors=k)
        X_r=lle.fit_transform(X)
        ax=fig.add_subplot(2,2,i+1)
        colors=((1,0,0),(0,1,0),(0,0,1),(0.5,0.5,0),(0,0.5,0.5),(0.5,0,0.5),
               (0.4,0.6,0),(0.6,0.4,0),(0,0.6,0.4),(0.5,0.3,0.2),)
        for label,color in zip(np.unique(y),colors):
            position=y==label
            ax.scatter(X_r[position],np.zeros_like(X_r[position]),label='target=%d'%label,color=color)
        ax.set_xlabel('X[0]')
        ax.set_ylabel('Y')
        ax.legend(loc='best')
        ax.set_title("k=%d"%k)
    plt.suptitle('LocallyLinearEmbedding')
    plt.show()

plot_LocallyLinearEmbedding_k_d1(X,y)