m0_47405013的博客

# 1、线性回归、逻辑回归、L1、L2正则化from sklearn.linear_model import LinearRegression,LogisticRegression,Ridge,Lasso#正则化# 2、神经网络from sklearn.neural_network import MLPClassifier,MLPRegressor# 3.支持向量机from sklearn.svm import SVR,SVC# 4.聚类算法：from sklearn.cluster impo

import numpy as npimport pandas as pdfrom sklearn.decomposition import PCA,TruncatedSVDfrom sklearn.linear_model import Lasso,Ridge,LinearRegressionfrom sklearn.pipeline import Pipelinefrom sklearn.preprocessing import StandardScaler,MinMaxScaler,Lab

1.单变量线性回归练习import numpy as npimport matplotlib.pyplot as pltimport pandas as pdimport warningswarnings.filterwarnings('ignore')# 1、导入相关包# 2、手写读取数据data=[[230.1,37.8,69.2,22.1],[44.5,39.3,45.1,10.4],[17.2,45.9,69.3,9.3], [151.5,41.3,58.5,18.5]

函数封装1.线性回归模型函数#模型def model(x,theta): return x.dot(theta2.代价函数#代价函数def cost(h,y): # return 1/(2*m)*(np.sum(h-y))**2 J=1/(2*m)*(h-y).T.dot(h-y) return J3.梯度下降函数#梯度下降函数def tidudecline(x,y,alpha,nums): m, n = x.shape #如果为单变量线

sigmoid函数sigmoid函数也叫Logistic函数，用于隐层神经元输出，取值范围为(0,1)，它可以将一个实数映射到(0,1)的区间，可以用来做二分类。在特征相差比较复杂或是相差不是特别大时效果比较好。Sigmoid作为激活函数有以下优缺点：优点：平滑、易于求导。缺点：激活函数计算量大，反向传播求误差梯度时，求导涉及除法；反向传播时，很容易就会出现梯度消失的情况，从而无法完成深层网络的训练。sigmoid函数是机器学习中的一个比较常用的函数，与之类似的还有softplus和softmax

神经网络

机器学习机器学习分为监督学习和无监督学习，监督学习为有标签，无监督学习无标签。监督学习分为分类和回归，而无监督学习包括聚类和降维。分类是预测的结果可列出，成为分类问题，为离散型；回归预测的结果是不可列出的，成为回归问题，为连续型。1.函数1.1 模型函数# 定义x*theta函数def model(x,theta): return x.dot(theta)1.2 sigmoid函数# 定义sigmoid函数 h(theta*x)=h(z)=1/(1+e-z)def sigmoid(

1. 神经网络底层import numpy as npimport matplotlib.pyplot as pltfrom sklearn.model_selection import train_test_splitfrom sklearn.metrics import confusion_matrix,classification_reportimport warningswarnings.filterwarnings('ignore')from sklearn.neural_netwo

from sklearn.datasets import load_iris,load_boston,load_breast_cancer,load_diabetesfrom sklearn.model_selection import train_test_splitfrom sklearn.preprocessing import MinMaxScaler,StandardScalerimport numpy as npimport matplotlib.pyplot as plt#加载数据集

支持向量机（support vector machine）1. 支持向量机练习1```pythonimport numpy as npimport matplotlib.pyplot as pltfrom sklearn.datasets import load_irisfrom sklearn.svm import SVC,SVRfrom sklearn.model_selection import train_test_splitimport warningswarnings.filt

1.聚类（k=4）from sklearn.cluster import KMeansimport numpy as npimport matplotlib.pyplot as pltimport warningswarnings.filterwarnings('ignore')from pylab import *mpl.rcParams['font.sans-serif'] = ['SimHei']# k-means# k均值聚类算法# 1.导入聚类包 from sklearn.cl

import pandas as pdimport numpy as npimport datetimeimport warningsimport matplotlib.pyplot as pltfrom sklearn.preprocessing import StandardScalerfrom sklearn.cluster import KMeans,DBSCANplt.rcParams['font.family']='SimHei'warnings.filterwarnings(

1. 图片读取import cv2import matplotlib.pyplot as pltos.environ['TF_CPP_MIN_LOG_LEVEL']='3'#读取图片img=cv2.imread(r'E:\ana\envs\tf14\day06\aaa.jpg')print(img.shape)#img.shape=(750, 500, 3)print(img.size)#1125000=750*500*3print(img.dtype)#uint8#显示图片cv2.im

1. 分布散点图、分布密度图、箱线图、小提琴图import seaborn as snsimport matplotlib.pyplot as plt#分布散点图def strip(): data=sns.load_dataset('tips') print(data) sns.stripplot( data=data, x='day', y='total_bill', hue='sex', do

sklearn机器学习算法1.逻辑回归： from sklearn.linear_model import logisticRegression2.朴素贝叶斯： from sklearn.naive_bayes import GaussianNB3.k-近邻：from sklearn.neighbors import KNeighborsClassifier4.决策树：from sklearn.tree import DecisionTreeClassifier5.聚类算法：from sklear

1. pandas独热import pandas as pddef main(): info_list=[ ['greed','A'], ['blue','B'], ['red', 'C'] ] df_info=pd.DataFrame( data=info_list, columns=['color','level'] ) print(df_info) df_1=pd.get

1.kmeans肘部图和DBSCAN轮廓图import pandas as pdimport numpy as npimport matplotlib.pyplot as pltfrom sklearn.cluster import KMeansfrom sklearn.metrics import silhouette_score# 3.读取drink.txt文件，完成一下处理data=pd.read_csv('drink.txt')print(data)# x=data.iloc[:,

from sklearn.feature_extraction.text import TfidfVectorizerfrom sklearn.decomposition import PCA,TruncatedSVDfrom sklearn.linear_model import Ridgefrom sklearn.preprocessing import StandardScaler,MinMaxScalerfrom sklearn.metrics import r2_score,roc_auc

import numpy as npimport matplotlib.pyplot as pltdef main(): np_datas_train=np.array( [[3,104,0], [2,100,0], [1,81,0], [101,10,1], [99,5,1], [98,2,1] ] ) test_x=np.array([

1. 泰坦尼克号练习import numpy as npimport pandas as pdimport matplotlib.pyplot as pltimport seaborn as snsfrom sklearn.preprocessing import StandardScalerfrom sklearn.model_selection import train_test_split,GridSearchCVfrom sklearn.linear_model import Lass

1.kmeans底层import numpy as npimport warningswarnings.filterwarnings('ignore')# 1.使用K-means底层，完成以下操作# 数据集[[1, 2], [2, 2], [6, 8], [7, 8]]x=np.array([[1, 2], [2, 2], [6, 8], [7, 8]])# (1)将数据集使用合适方式进行处理（10分）# (2)使用随机方式，初始化质心（10分）center=np.array([[1,2]

1.贝叶斯调库# 1. 导入必要的数据集（10分）import numpyfrom sklearn.datasets import load_irisfrom sklearn.model_selection import train_test_splitfrom sklearn.naive_bayes import GaussianNB # 高斯朴素贝叶斯from sklearn.metrics import accuracy_scorefrom sklearn.metrics import

1.聚类底层import numpy as npimport pandas as pdimport matplotlib.pyplot as pltimport warningsfrom sklearn.neighbors import KNeighborsRegressor,KNeighborsClassifierwarnings.filterwarnings('ignore')plt.rcParams['font.family']='SimHei'def main(): #1

1.轮廓系数import pandas as pdimport numpy as npfrom sklearn.cluster import KMeans,DBSCANimport matplotlib.pyplot as pltfrom sklearn.metrics import silhouette_scoredef main(): np_datas = np.loadtxt(r"./good.txt", delimiter=", ") x = np_datas[:, :-

1.PCA降维底层方法一：import numpy as npimport pandas as pdimport matplotlib.pyplot as pltimport seaborn as snsdef pca(n_components,data): # print(data) x = np.cov(data) # print(x) # 取出特征值和特征向量 t_value, t_vector = np.linalg.eig(x) # p

知识点总结1.机器学习知识框架**分类（classification）????*支持向量机（svm）、逻辑回归、最近邻、随机森林、Naive Bayes、神经网络。回归（regression）:线性回归(linear regression)、岭回归(ridge regression)、支持向量回归（svr）聚类（clustering）:K-Means、Spectral Clustering(谱聚类)、均值漂移（mean-shift）降维（demensionality reduction）:主成分分

import datetimeimport numpy as npimport pandas as pdimport seaborn as snsimport matplotlib.pyplot as pltfrom sklearn.metrics import roc_curve,r2_score,roc_auc_scorefrom sklearn.preprocessing import StandardScalerfrom sklearn.decomposition import PCA

1. 结巴words='天王盖地虎我在山中走'd=jieba.cut(words,cut_all=False)dd=list(d)print(dd)print(' '.join(dd))print('--------------------------')d1 = jieba.cut(words, cut_all=True)dd1=list(d1)print(dd1)print(' '.join(dd1))print('------------------------------')

1. seaborn绘图import seaborn as snsimport numpy as npimport matplotlib.pyplot as pltfrom sklearn.datasets import load_bostonfrom sklearn.decomposition import PCAfrom sklearn.model_selection import GridSearchCV,train_test_splitfrom sklearn.preprocessin